Traditional technologies

Some technologies are referred to as `traditional' because farmers have used them for many years, in some cases for centuries. Underhill (1984) first suggested a classification of traditional technologies, which is used here with some modifications, Table 1.

Table 1: A physical classification of traditional smallholder irrigation

Classification	Description	Some examples
Water harvesting	Involves making better use of natural run-off. Many approaches. Collecting run-off from a catchment and concentrating water in a smaller cultivated area. Large schemes, several hectares of catchment, small schemes may involve individual plants. Success depends on having right ratio of catchment to cropped area. New developments applying chemicals, plastics to catchment to increase run-off.	Ancient agricultural areas, Negev Desert, Israel. Pasture improvement by flood water retention, Niger. Micro-catchments for fruit trees, Botswana rice cultivation, Gambia.
Swamp irrigation	Fresh water swamps protected from saline seawater by bunds/dykes. Used for growing rice. Also tidal swamps planted after rains leach soil.	Gambia River, Gambia.
	Inland valley swamps - small valleys (1-100 ha) where season/perennial streams can be used/controlled for paddy rice cultivation.	Sierra Leone, Burundi
	Bolilands and dambos - depressions in swamp grasslands.	Rokel River, Sierra Leone, Guinea, Mali, Burkina Faso, Côte d'Ivoire, Ghana, Malawi and Zambia.
	Lake swamps - large areas of flat plain flooded as lake level rises.	Lake Victoria, Tanzania
Spate irrigation	Water spreading - spreading floodwater in rivers and wadis across cultivated land in a controlled manner.	Lower Omo Valley, Ethiopia, food and fodder crops
River flood plain irrigation
- Wet season	Flooded lands. Usually rice grown in floodwaters. Many techniques.
	Controlled irrigation - controlling the rise in floodwater using dykes, canals and sluice gates. Used for both deep water and paddy rice.	Niger and Bani Rivers, Mali.
- Dry season	Recessional irrigation. Impounding receding floodwaters with earth bunds.	Rice cultivation along most large rivers in West Africa.
	Residual moisture. Similar to recessional, water is stored in soil rather than on the surface.	Dry season cultivation of dambos or vleis in Malawi, Zambia, and Zimbabwe.
	Pumped irrigation. Large areas of flood plain where surface water storage and shallow ground water can be exploited using lifting devices, often for vegetables.	Increasingly common in Nigeria also along most large river flood plains. River Omo, Ethiopia
	Shadufs - traditional technique of lifting small quantities of water using the `lever principle'. Usually low lift 1-2 m.	Most large river flood plains in West Africa, Niger, Bani Senegal, common along Nile. Tana River, Kenya.
Hill irrigation	Land irrigated some distance from water source, supplied by canal or pipe. Source may be a stream, small dam storage, gravity or pumped.	Traditional cultivation by the Chagga tribe on the slopes of Mount Kilimanjaro, Tanzania, Malawi, Ethiopia
Groundwater irrigation	Involves exploitation of groundwater down to 15 m.	Exploited for small gardens, e.g. Burkina Faso, Niger, Togo, Benin, Zimbabwe, Botswana

Brown et al. (1995) attempted to show how these technologies where linked to farmer control, i.e. the level of control that a farmer would have over the design and investment decisions and over water management, Figure 1.

Figure 1: Farmer control of irrigation schemes (Brown et al 1995)

Water harvesting

Water harvesting is a method that has been around for centuries but is not widely exploited in sub-Saharan Africa. It is ideally suited to arid and semi-arid areas where rain-fed crops cannot be grown with any certainty because the rainfall is both unreliable and highly variable (FAO 1994). The rainy season is often short with no assurance of when it will start and finish and there may be frequent long dry spells. Rainfall is collected from surrounding areas and channelled as run-off onto farms to add to the rain that falls directly onto the crops. Interestingly, no one doubts the critical importance of rainfall. Although few policy-makers recognize the importance of run-off, which is the inevitable product of excessive rain. It is a curious paradox that farmers recognize and exploit the natural concentration of rainwater in valley bottoms and local depressions, yet the overriding perception driving policy is that run-off is a hazard. This view is fuelled by the prominence given to the concerns about soil erosion, which is one product of run-off and has been the focus of research and extension efforts to curb it since the 1930s. Water harvesting recognizes the potential value of run-off as a resource and aims to control the process in order to mitigate the hazard.

Run-off is collected from a large catchment area and channelled to increase the water available in a smaller growing area. There are micro-catchment systems, which are modest in size, where water is collected from land adjacent to the farm and channelled directly onto the fields. On an even smaller scale micro-catchments can be constructed around individual plants (often trees). There are macro-catchment systems that have large water collecting areas; often a considerable distance from the farming areas and can serve many farms. Substantial quantities of water can be collected from barren and fallow land and channelled into the cropped fields.

Water harvesting in Tanzania In Tanzania water harvesting is used successfully on a large scale for growing rice (Hatibu 1999). Indian migrants are believed to have introduced the so-called majaluba rainfed rice system in the 1920s. It is used in the lowlands of Tanzania where seasonal rainfall can be as much as 600-900 mm and runoff naturally collects in the valley bottoms making it ideal for paddy rice. Its adoption was not led by external change agents and was not fostered by external subsidies, but nevertheless it has spread steadily since the 1930s. Official data now show that the majaluba systems contribute 35 percent of total rice production in Tanzania. Recent research has been carried out to try and introduce micro systems in the drier areas for maize production. The idea is that micro systems would give smallholders more control over their farms. However, when they were invited to evaluate the micro-catchment trials, farmers understood the benefits of rainwater harvesting but were reluctant to adopt the system. They were more interested in the greater potential of using macro-catchment systems and argued in favour of more ambitious attempts to harvest runoff on a larger scale. So far the limited trials with macro systems for maize are mixed. Proper control over distribution of harvested runoff within the cropped area is more problematic for deficit-irrigated crops than is the case with majaluba rice systems. There is also clear evidence that failure to provide proper control over the distribution of runoff can lead to serious erosion. Too much water can be as big a problem as too little. The need for cooperative group action can also give rise to disputes over water sharing. So whether farmers will continue to prefer macro-systems to micro-systems as they acquire more experience in using them for maize production remains to be seen. However, one significant outcome of the research is that government sees runoff as a benefit rather than just a hazard and the cause of soil erosion. Development of rainwater harvesting is to be included in the Tanzania National Water Resources Management Policy.

In sub-Saharan Africa, although there are examples of water harvesting, it occupies a neglected middle ground between soil and water conservation and irrigated agriculture. Both these extremes have received far greater attention.

Swamp irrigation

Swamps have been traditionally used for irrigation throughout the region. In mangroves and coastal swamps earth banks or bunds are built to exclude seawater and allow fresh water from inland to enter the cropped area, which is then planted with rice. The main areas for mangrove and coastal swamp irrigation stretch from The Gambia round to Liberia.

Inland valley swamps are naturally occurring swampy valley bottoms (up to 100 ha) used to grow rice. Small weirs and channels are constructed to divert natural drainage water around a swamp. The area is then divided into paddy fields, sluices are used to control the flow of water into the fields and a central drain allows excess water to escape back into the natural drainage (FAO 1984). This is very common in the humid parts of West Africa such as Sierra Leone and Liberia and in Burundi.

In the wet season many of the African lakes flood large areas of flat plains and these are used for rice cultivation. Examples of this type of irrigation exist in Tanzania beside Lake Victoria and Lake Malawi, in Malawi by Lake Chilwa and in the countries bordering Lake Chad. Also important are the traditional irrigation practices found in bolilands and dambos. These are low, saucer shaped depressions in swamp grasslands such as those associated with the Rokel River in Sierra Leone. There are similar areas in Burkina Faso, Ghana, Guinea, Côte d'Ivoire, Mali, Malawi and Zambia.

Spate irrigation

This is a method of water spreading usually from flash floods in wadis (FAO 1987), which is a common practice in some arid countries such as Saudi Arabia and Yemen. It is not commonly practised in sub-Saharan Africa, although there are instances of its use in the Lower Omo Valley in Ethiopia for both food and fodder crops. The technology requirements need not be great though considerable skill is required to successfully divert and spread the water.

Flood plain irrigation

Flood plains are principally used for growing rice during the wet season. There are a variety of techniques involving different amounts of water control that overlap to a greater or lesser degree. Some of the more distinct types of wet season irrigation include:

• Deep-water rice grown on the flood plains of the Niger and Bani rivers in Mali. Holdings are 1.5 ha with earth dykes constructed to prevent the fields being flooded whilst the rice plants are still immature. Problems arise if the floods arrive too late or too early or if they rise too fast and submerge the rice. Because of the high risk of failure, this technique tends to be secondary to rainfed subsistence crops cultivated above flood plain level.
• Flood irrigation involving the control of floodwaters to 50 mm/day for deep-water rice and 30 mm/day for paddy rice. Various levels of technology are possible from simple earth banks to dykes, canals and sluice gates.

In the dry season, shallow groundwater or open water left after the floods is used through a variety of gravity and water lifting techniques for:

• Recessional irrigation. Areas recently inundated by floods are utilised by impounding the receding water with small earth bunds. This type of irrigation is found in all the major flood plains in West Africa for rice irrigation and close to lakes (e.g. Lake Chad) for growing maize and sorghum.
• Residual moisture. This is similar to recessional irrigation but utilises the water stored in the soil rather than water retained on the surface. It is a very common form of agriculture on all major flood plains. In East Africa it is found on many small dambos or vleis (seasonally wet areas in small catchments).
• Pumped irrigation. Pumping from rivers, streams and lakes is a simple activity and can usually be managed on an individual basis. The additional cost of pumping usually means that farmers are growing high value crops for sale in markets rather than low value subsistence crops.

The shallow depth aquifers that are scattered over more than 100 000 ha of Niger as well as northern Nigeria, Mali, and other Sahelian countries, also provide possibilities for successful smallholder schemes. The aquifers must be rechargeable (as they are in Niger and Nigeria). An obvious limitation to the expansion of irrigation on this basis is the limited area where such aquifers exist, and the possibility of overtaxing the recharging capacity of these areas. A wide variety of water lifting methods are used covering a range of traditional types such as archimedian screws and the shaduf to the modern petrol and diesel driven pumps. There has been a growing interest in treadle pumps throughout sub-Saharan Africa since their introduction from Bangladesh and India (Kay and Brabben 2000). Generally motorized pumps are most common in the richer countries that have ready access to spare parts and fuel, e.g. Nigeria. In water scarce areas there is also a growing interest in the use of modern distribution

Individual pump schemes in Nigeria Farmers in northern Nigeria lost their traditional use of the fadamas along the rivers following the construction of dams to control the river floods for urban water supply and irrigation. As an alternative they turned to small-scale irrigation using shallow groundwater recharged by the river and lifting it by shadouf or calabash in the dry season to grow vegetables for local and city markets. In the early 1970s a few farmers, with help from relatives, bought small pumps from private traders. In 1982-83 an agricultural development programme based in Kano sold over 2 000 pumps for cash to individuals or small farmer groups. Engineers introduced low-cost well technologies from India, which reduced well construction by two thirds with a commensurate increased return on tubewell investment. This has been one of the most successful irrigation developments in Nigeria with many thousands of pumps being used by private farmers. Maintenance is well established and so farmers have confidence in the technology. However, external monitoring was necessary to avoid depletion of the aquifer. Interestingly in the 1980s some farmers started to grow wheat on the fadamas in response to the high wheat prices in the country. A useful example of the way in which private farmers can and will respond to the market if the price is right.

equipment such as trickle and sprinkle irrigation for cash cropping.

Hill irrigation

There are many small, irrigated areas that are situated some distance from their water source and supplied by mainly open channels and low-pressure pipe systems. The water source may be a stream or a small dam supplying water by gravity or use of some type of simple lifting equipment. An individual farmer or a group of farmers working together to share the water resource can practise this type of irrigation. Such systems are common in the hilly areas of central Ethiopia, Tanzania and Malawi.

Groundwater irrigation from medium to deep aquifers

Medium to deep aquifers (>30 m) can be exploited all year and usually require more sophisticated pumping than is commonly found in traditionally irrigated areas. Usually submersible pumps driven by electric motors or diesel engines are needed. There are numerous examples of this type of water lifting and the agriculture based upon it throughout Africa and other parts of the world, notably Bangladesh, Pakistan and India. This type of irrigation is only feasible for cash cropping because of the high cost of well construction and pumping or when it is subsidised in some way.

Communal scheme experiences

All the methods described above can be practised by individual farmers on their own or by several farmers working together and sharing a common water supply. Once farmers group together the management of the supply system needs to be re-examined. If the scheme is small enough, the farmers may choose to operate it themselves. As the size increases a manager may need to be hired full time to undertake management of the irrigation scheme. Either way there is a significant shift in the management of the irrigation from an individual who can irrigate as and when they please to a group who must work to an agreed schedule of sharing in terms of quantity and timing.

This communal approach is most favoured by aid agencies and national governments because it is seen as a way to help many farmers at the same time. Agencies often step in, with the best of intentions, to satisfy a need or to act as a catalyst for development. In so doing the aid agency or government brings its own set of rules, which may not be compatible with what is required on the ground. Agencies need to disperse funds in line with their targets and are often constrained by time. The temptation to speed the process of development is always present and, as a result, so is the risk of killing the very thing that the agency is trying to foster by pouring in too much support too quickly. If the community cannot absorb the funds fast enough and do not have the institutional framework, farmers may quickly lose any sense of ownership of the project as they rely on a culture of handouts and support from outside. In such situations the project has a high risk of failure. The challenge is to provide the correct amount of support over the right time period to foster successful development. Underhill (private communication) once said that if you `interfere' in a privately initiated irrigation development by injecting more than 10 percent of the total cost then you are in danger of killing it off. This is not an accurate figure but it does give some order of magnitude of what can be done.

There are many examples of communal schemes in sub-Saharan Africa ranging from the very large (e.g. schemes in Nigeria, Kenya) to the more modest (e.g. Senegal, Zimbabwe, Tanzania). Most have run into serious problems, although there are some successes. They tend to be top-down in approach and the technologies used are not always conducive to simple water management practices and often add to the problems of sharing available water resources rather than helping to solve them.

For example, open channels are the most common system of supply and these are usually fitted with upstream control structures. This is one of the most complicated systems of water management available and so it is little wonder that farmers and scheme managers find it difficult to distribute water properly. There is no easy way of knowing if too little or too much water has been put into the canals without some sophisticated water indenting procedures and flow measuring devices. The hydraulic structures can favour those at the top end of the canals and there is usually a constant need to adjust the control gates. In contrast, pipe supply systems are much less common but are much simpler to operate and respond rapidly to changes in water demand. They can be less wasteful of water because it is easier to turn off a pipe supply when it is not needed. It is not possible to turn off a canal without draining it down.

When looking to the future it is vitally important to review past experience of similar developments to see what lessons can be learned about what and what not to do. Several comprehensive reviews of the Sahel region have been made in recent years (e.g. Moris 1984 and 1987, Brown and Nooter 1992). The two types of development are recognized in the reviews: those in which farmers have funded irrigation themselves using their own resources and those that have received external support from an agency, either in the form of direct aid or through a special loan arrangement. The former are usually based on the farmers' perceptions of family needs and those of the local markets. Investment is likely to be made with great care to minimize the risk of failure. The latter have a much higher risk of failure for a range of reasons, and the fact that they are for smallholders (as opposed to large-scale irrigation) is no guarantee of success.

Review of the Sahel experience

Both Moris and Brown and Nooter's studies concentrated on the Sahel region. The authors found that there were some strikingly successful privately operated smallholder schemes in Mali, Mauritania, Niger and northern Nigeria as well as in Burkina Faso, Chad and Senegal. In some cases, these developments were entirely spontaneous, and in others, they were supported by NGOs or with minimal government assistance. Some farmers took advantage of earlier investment and used the infrastructure from earlier, failed large-scale projects. In some cases, expansion was facilitated by the use of new low-cost construction techniques that reduced the cost of installing tubewells and hence the costs of irrigation.

The reasons for success varied from country to country and depended on a range of technical and socio-economic circumstances. For example, in Niger private sector development had expanded well and responded to market forces. In Nigeria there was a similar story but it had been helped along by subsidised prices for equipment and farm inputs. Improved rural roads enabled farmers to market their surplus production and gain access to traders and decentralized food-processing plants. Mali demonstrated the extremes of success and failure. Where there is success in the private smallholder sector and failure in public sector large-scale irrigation. The Office du Niger, a parastatal corporation, concentrated to such an extent on `public services' that it eventually had two staff members for every three farmers and for every 11 ha of irrigation. Only 20 percent of farmers' fees went for actual inputs. Chad, although suffering from civil strife for more than a decade, has seen success built on simple low-cost techniques funded by non-governmental organizations.

Senegal exemplified the changing focus of irrigation. This shifted from large, publicly managed irrigation systems with the farmer as labourer, to unsuccessful attempts to make parastatal irrigation agencies more efficient, to government assisted smallholder irrigation, to experimenting with non-public sector irrigation. The early phases of this experience suffered from the inappropriate construction of irrigation perimeters and from selecting crops that were not economically viable. Project design, operations and reform attempts by both the Government and donors were carried out from the top-down, ensuring a low level of farmer enthusiasm. An attempt by the Government to stimulate smallholder irrigation failed when the plot size determined by the Government was too small to encourage farmer participation.

Successful projects in Burkina Faso have been the result of a relatively encouraging macro-economic framework. Farmers have often been able to use the irrigation infrastructure from "failed" projects and the migration of men looking for work made irrigation a profitable and necessary channel of agricultural development for women.

Cameroon has been an exception. Large-scale irrigation has been successful but mainly because of strong expatriate management using farmers as labourers rather than decision-maker. Farmers have accepted this situation because of the high financial returns and because of the lack of other opportunities.

From their comprehensive review, Brown and Nooter (1992) suggested several characteristics that are common to successful schemes:

• Technology is usually simple and low-cost (most frequently small pumps drawing water from shallow aquifers or rivers and streams).
• The institutional arrangements are private and individual.
• The supporting infrastructure permits access to inputs and to markets for the sale of surplus production.
• There is a high financial (cash) return to farmers at the times when they most need cash.
• Farmers are active and committed participants in project design and implementation.

The authors found the perception of `success' was different for irrigation technicians, donors, governments and farmers. Their view was that success could only be assessed in terms of the farmer's definition.

The most effective arrangements for schemes that were larger than individual ownership were found to be (in decreasing order of success): extended family groups; private voluntary groups; water users' associations and then cooperatives. They concluded that project design should be based on the following concepts:

• Encourage smallholder, private and individual investment in irrigation and implement projects with many small components, where NGOs frequently are effective implementing intermediaries.
• Employ methods that ensure early and full farmer participation in project design and operations, devoting as much staffing and funding to studying the farmers' economic and social situation.
• Identify and disseminate low-cost small pump technology and tubewell construction techniques and include provision of soil and water surveys beyond the purview of individual farmers but essential for sustainable irrigation.
• Ensure a macro-economic framework that will reward the farmers' and merchants' desires for low-costs and high returns based on real costs. Both financial and economic rates of return must be satisfactory on a sustained basis and avoid unsustainable subsidies, since their removal could lead to a project's collapse.
• Ensure the availability of foreign exchange needed to secure supplies of spare parts, inputs and adequate repair facilities and ensure that there is the essential infrastructure needed for access to imports and markets.
• Make provision for training in maintenance and irrigation techniques for farmers, preferably through private sector suppliers; and ensure that environmental considerations are provided for, such as the effect of irrigation on health, drainage and erosion control, which may not be evident to the farmer.

Role of traditional technologies in sub-Saharan Africa

A wide range of established and well-documented traditional technologies is available for use by smallholders. These clearly play a significant role in the region and will continue to do so. There is still considerable room to adapt traditional technologies to different circumstances, which because of their low cost and simplicity can be used and maintained by smallholders with little or no external support. They are particularly suited to subsistence farming and equally have an important role to play in the transition to cash cropping. For this reason they have attracted much interest from aid donors and governments wishing to support subsistence farming. Many lessons learned, about what needs to be done, have been concerned more with the ways in which such technologies can be introduced into farming systems rather than the technologies per se. However, there are still concerns about the technology and the ability of people to design and construct good engineering works that will provide lasting service. There is also a danger that `low-cost' solutions, which may be attractive to aid donors, may become a euphemism for poor engineering.