Some innovations don’t wait for experts

A report on applied research

by Bangladeshi peasants

If necessity is the mother of invention, then one should certainly expect to find much evidence of farmers' ingenuity in Bangladesh. That country has a population of around 90 million crowded into a gross area of 150 000 km2 of which about 8.8 million hectares are considered agricultural land. Ninety percent of the country's population is classified as rural.

To visitors coming from countries where single-crop cultivation is the norm, Bangladesh's agriculture does, indeed, often I seem ingenious. Double- and triple-cropping practices are common. Two of the most widespread practices are broadcast "upland" rice sown with premonsoon showers, followed by a second rice crop transplanted in the middle of the monsoon season; and, on land that is flooded somewhat deeper during the monsoon season, a mixture of a quick-maturing upland rice and a longer maturing deep water rice, sown together on premonsoon rainfall but harvested during and after the monsoon season respectively, and followed by a dry-season crop of pulses, oilseeds or wheat grown on residual soil moisture.

In a climate permitting plant growth throughout the year-in areas where monsoon-season flooding is not too deep and the soils store sufficient moisture for dry-season use (or where they can be irrigated) -one should perhaps not be surprised that multiple cropping and intercropping practices are so common. But what also amazes the visitor is the yew, fine adjustment of the farmers' crops and cropping patterns to the specific micro-environment of their fields. A few inches' difference in elevation in relation to expected flooding depth and duration can cause farmers to plant different rice varieties. Over time, this has led them to select varieties suited to their specific situations. The result is a wide diversity of rice varieties: to date, the Bangladesh Rice Research Institute (BRRI) has identified about 4 500 different varieties (and considers that this may be perhaps only half the actual number in use).

A regrettable ignorance

This process of varietal selection is still going on. The Director of BRRI says he knows of three cases where farmers have made their own selections from JR-8, one of the earliest high-yielding dwarf varieties released by the International Rice Research Institute (IRRI). It is of interest to note that, in each case, these selections were made for greater plant height. This reflects the farmers' perception that the new dwarf varieties were not well adapted to situations where full water control could not be provided. To be fair, BRRI plant breeders also quickly recognized the need for greater seedling length and plant height, and included these plant characteristics in their breeding objectives from 1970 onward.

The above example well illustrates the fact that farmers-as a community-are not as dependent on official research and extension programmes governments and aid donors commonly seem to assume. One reason for this is that official research and extension activities are dependent on the size of government fund allocations and, historically, there have been periods when these allocations were inadequate to carry on these activities on more than a care-and-maintenance basis. Another reason is that government research stations-even where land shaping, tractor cultivation, liberal irrigation and mono-crop research practices have not made conditions irrelevant for farmers' real-world situations-cannot possibly represent the whole range of agroecological conditions existing in a countm or a region. Of course, the latter constraint can be overcome, in part, by on-farm trials, but even these trials rarely cover more than a small proportion of the micro-agroecological conditions under which a country's farmers actually grow their crops. The final stage of adaptive research must inevitably be left to the farmers, therefore.

However, this article does not aim to describe this aspect of adaptive research. Rather, it aims to illustrate more basic adaptive research carried out by the farmers themselves, research of which government and extension officials, as well as policy-makers and planners, often seem to be unaware, both in terms of results and as an activity. This ignorance is regrettable on two main grounds: the waste it implies of a tremendous resource of native talent and information which officials could use to amplify and accelerate their own research and development activities; and the waste of the opportunity for dialogue between professionals and practitioners who have - although perhaps unrecognized or unformulated - a mutual interest in finding solutions to agronomic problems.

The following illustrates a number of farmer innovations I observed during my travels in Bangladesh. The list is not complete: each field trip seems to reveal a new example. Also, each successive year that I travel indicates the rapid spread of successful innovations. The impression is gained of an unofficial research and extension network operating independently-even obliviously - of government programmes, often more practically oriented than the latter and, because of this, apparently more successful in terms of new adopters. These comments are not meant to be disparaging of government research and extension activities as such. Obviously, there is a need for official R & D programmer. But what I would like to underline is the lack of official attention-because of lack of awareness-that is paid to such farmer activities. The inevitable consequence of this neglect is a retarded rate of agricultural development.

As a mulch

Farmers in different parts of Bangladesh have developed many multiple-cropping practices. I once counted as many as fifteen different vegetables and spices growing intermixed in two adjoining plots that together occupied less than 0.04 ha. However, one specific example of intensive multiple cropping is particularly interesting. This example occurs in the Chandpur-Daudkandi area of Comilla District in the east of the country, on land that is deeply flooded for 4-5 months in the rainy season.

Because of deep flooding-normally 1-3 m deep at its peak in July-Au gust - deep-water rice is the main crop grown in the rainy season. Before this crop is harvested in November or December, some farmers broadcast seed of mustard or fodder legumes (Phaseolus mango and Lathyrus satires) on the wet soil surface. When the rice crop is ripe, the heads are harvested by hand, leaving the tumbled straw on the fields as a mulch. The mustard is harvested and the fodder crops are grazed off in late January or early February. The rice straw is then removed or burned on the field, and the soil is immediately ploughed.

No textbook solutions

At this time, three months after the end of the rainy season, the topsoil of the deep estuarine silts is still moist. Before the weather begins to warm up in mid-February, the farmer broadcasts sesame or sows Italian millet (Setaria italica), or he transplants chill) seedlings. With the first heavy shower in March or April, he hand-weeds these crops and at the same time undersows them with his next deep-water rice crop (which some farmers also intermix with a quick-maturing upland rice or with jute). In May or early June, before the floodwater rises, the farmer harvests his millet, sesame or chill) crop by pulling up the whole plants. The deep-water rice continues to grow with the rising floodwater; if upland rice or jute has been intermixed, it is harvested in July, before the floodwater becomes too deep.

This example provides a number of interesting lessons which show that the farmers' response to challenging situations is often different from the textbook solutions.

Intensive cropping practices do not necessarily require mechanized cultivation. Even though three or four crops are grown in a year in the example, the land is ploughed only once.

The use of modern high-yielding varieties is not the only way in which farmers can increase crop production. In the Chandpur-Daudkandi area, it appears that the sequence of intensification has been:

· initially, a single crop of deep-water rice;

· then, the admixture of a quick maturing upland rice (except in depression sites);

· later, the addition of an early dryseason crop of pulses or oilseeds sown through the deep-water rice crop; and

· finally, the addition of a second dry season crop, through which the deep water rice crop is intersown.

Irrigation is not essential for dryseason cropping on deep loamy soils with a high moisture-holding capacity.

The use of fertilizers is not essential for continuous cropping on seasonally flooded soils. All the crops grown in the Chandpur-Daudkandi area are, in practice, low input crops. Soil fertility is maintained partly by the use of pulses in the rotation, but probably mainly by the seasonal flooding. The latter provides biological nitrogen by way of blue-green algae. Also, the seasonal fluctuation between aerobic and anaerobic soil conditions makes phosphorus and possibily potash more readily available than would be the case under continuous dry-land conditions. (Soil fertility is not maintained by annual accretions of silt. Most of the Chandpur-Daudkandi area is flooded with ponded rainwater, not with river water. The farmer may not be aware of this distinction, but he is aware that, in practice, seasonal flooding maintains the fertility- actually, the productivity-of his soil.)

Intensive cropping practices do not necessarily require or create more employment. The three - or four-crop rotation described probably requires little more labour than the initial single crop of deep-water rice: there is still only one ploughing; and weeding costs are probably no higher (and may even be lower) because the second dry-season crop helps to shade out weeds in the deep-water rice crop, which normally requires several weedings, and the harvesting of this dry-season crop by uprooting has the effect of an intertillage operation. Also, the sowing and harvesting of the first and second dryseason crops are done outside the periods of peak labour demand. The farmer therefore is able to optimize the use of his family labour.

Such intensive practices are possible only on superior soils, of course. Yet, even on problem soils, farmers have found novel ways of intensifying crop production, which would probably not have been thought of by orthodox research scientists. Two examples of this can be cited: the dibble-sowing of rice on saline soils and the cultivation of wheat on unsuitable rice soils.

Soils in the coastal areas of Noakhali District in the southeast of Bangladesh become saline at the surface during the hot weather preceding the monsoon season. Traditionally, these soils have been used for a single crop of transplanted rice grown in the monsoon season when the soils are not saline. In recent years, however, some farmers have also started to grow an upland rice crop in the pre-monsoon season. This they do by dibbling, that is, making planting holes about 5 cm deep with a strong stick and dropping a few seeds in each hole. Sowing is done late in February, after the end of the cool winter period, at which time the topsoils are still moist. It is interesting to note that the planting holes are left open: the farmers' idea, presumably, is to prevent a saline crust forming on the soil surface around the growing plants. Some farmers have tried sowing high-yielding rice varieties, but most use traditional varieties, which they consider more suitable for the environment.

From an orthodox scientist

According to the farmers to whom I spoke, dibble-sowing started 30-40 years ago. It has spread rapidly in the last years, not only across the broad Meghna estuary to Barisal District, but also to adjoining areas of Moakhali District where the soils are not saline. The practice is of interest in that it indicates a difference in farmers' perception and response from that which might have been expected from an orthodox scientist. The innovative farmer who first tried the technique presumably sought to place the seed below the surface salt crust. A scientist would probably have sought to remove the salt, either by irrigation (which is not available in the area) or by deep cultivation to prevent capillary movement of salt to the surface.

Dibble-sowing also has additional advantages. It is not necessary to prepare a fine seed-bed such as broadcast sowing requires. The seed is sown early, on residual soil moisture, without waiting for pre-monsoon showers to fall in March-April; this enables the quick-maturing crop to be harvested before the monsoon season rather than during it. The seed is placed in contact with moist soil, thus ensuring better germination than is obtained from the traditional practice of broadcast-sowing. Less seed is used than in broadcast sowing; the plants are uniformly spaced, providing higher yields; and weeding and harvesting are made easier than with broadcast-sown crops.

More labour is used in this practice than in broadcast-sowing. However, family labour can be used since there is not the same time pressure for completion of the operation in February as there is in the pre-monsoon season when land preparation and broadcast sowing may have to be completed within a few days while the soil is in a suitable moisture condition after rain, and there is competition with raking and weeding operations. Dibbling, in fact, is a suitable occupation for old men, and the sowing is often done by young children who do not have so far to bend as adults for placing seed in the holes.

Horticultural techniques

Wheat cultivation on the Barind tract soils of Bogra District in the northwest of Bangladesh provides an example of farmers jumping ahead of the scientists. Being a soil scientist, I had recommended that these impervious soils, puddled for transplanted rice cultivation in the monsoon season, were unsuitable for wheat cultivation in the following dry season. Three years ago, however, I found some farmers growing wheat on these soils. Not, of course, in the way in which wheat normally is grown, broadcast on the flat. These innovative farmers had made ridges by hand, as for cultivating potatoes-which are also grown in the area-and had sown two closely planted rows of wheat on each ridge. Also, they were irrigating the crop from dug wells or tanks (excavated ponds), applying frequent, small amounts of water down the furrows so as to avoid waterlogging the soil. The crop growth appeared excellent, a view obviously shared by neighbouring farmers, because the practice has spread widely during the following two years.

In retrospect, I recognized that cultivation on ridges and the application of frequent small amounts of water are the solution for cultivating dry-land crops on impervious soils, but I had not considered that the farmers would use horticultural techniques for cultivating a crop such as wheat. These soils now have to be classified as "conditionally suitable" for wheat, the conditions being that the land be free from the previous rice crop in time to sow wheat before mid-December, that ridges be made to provide an aerated cultivation layer, and that frequent, small applications of irrigation water be given.

The last example illustrates how farmers' innovations can upset planners' projections of the output expected from proposed investments. Hand tube-wells have come into widespread use for dry-season irrigation in several parts of Bangladesh. Conventionally, it has been assumed that one hand pump can irrigate 0.16 ha of transplanted rice and up to 0.4 ha of wheat. (In practice, many pumps irrigate less than these areas, because most farmers' fields occupy less than 0.08 ha.)

However, travelling in Kushtia District in western Bangladesh last winter, I stopped to photograph a hand pump being operated by foot - itself an innovation-and discovered in conversation with the pump owner that he was irrigating 2 ha of tobacco from his pump. After irrigating one field, the farmer would lift the two pipes and resink them in another field - the farmer's holding, like that of most Bangladeshi farmers was fragmented -then irrigate that field, and so on until he had irrigated all his fields in rotation. In the soft alluvial soils of Bangladesh, where the water table is usually close to the surface, at least during the cool winter months, it is a job of no more than a couple of hours to lift, move and resink two or three pipes and attach a hand pump. Subsequently, I learned of farmers using a single hand pump to irrigate more than 2 ha of dry-season crops.

Enormous implications

Not all Bangladeshi farmers own 2 ha of land, of course. However, the fact that a few pioneer farmers have shown that such an area can be irrigated by one pump indicates that planners, aid donors and extension workers must drastically revise their previous projections regarding the contribution which these small pumps can make to irrigated crop production: the difference between 0.2 ha and 2 ha is tenfold! On their own initiative, some farmers are now renting out their pumps to other farmers. One voluntary agency, too, is providing landless farmer groups with pumps on hi repurchase terms, so that a mobile, labour-intensive irrigation service is given to cultivating landowners and sharecroppers. In these days of rapidly rising fossil fuel costs, the implications of these innovative irrigation practices are enormous. If one hand pump costing less than $100 can irrigate 2 ha (of dry-land crops, not transplanted rice), it is a much more economical investment than a shallow tube-well costing more than $ 1000 and irrigating 4-6 ha, or a deep tube-well costing $10 000 and irrigating 24-40 ha. There are considerable social benefits, too, in terms of employment provided.

It is imperative that R & D officials become aware of farmers' research efforts and make use of them. Extension could use innovative farmers' fields as demonstration sites, and could organize demonstrations of the new practices in other areas where conditions are similar. This accelerated pace of research and extension could lead to a more rapid increase in agricultural production. Inevitably, too, the spinoff in enhanced mutual respect engendered between officials and farmers would produce additional benefits in production through greater career interest and farmers' cooperation.