How fast can smallholder irrigation develop?

It is important that agencies wishing to support irrigation development have some idea of how fast future irrigation development can take place so that the agencies can plan for the labour and technical resources they will need to provide. Given the current socio-economic constraints on smallholders the question is a difficult one to answer. Removing constraints can encourage farmers to move into irrigation and speed up the process but this is not easy and so major changes are unlikely in the foreseeable future.

One option is to look at what has happened in the past decade as an indicator of what might happen in the future.

Review of experience from sub-Saharan Africa

A recent review of IFAD's smallholder projects (Internal review 2000) is of particular interest not only because of its wide geographical spread but because it highlights the challenges that face any international agency wishing to support smallholder irrigation in what is essentially the private sector. As the majority of such investments tend to go to the larger schemes involving communities, rather than to support individuals, they have inherently higher risks built into them. Past experience has shown that individual farmers developing irrigation on their own or with family have a much lower risk of failure than those who work in groups or communities requiring collective responsibility and a sharing of the assets.

In 1986 IFAD launched its special programme for countries affected by drought and desertification in sub-Saharan African in response to the devastating droughts of the 1980s with a sub-programme: Small-scale Irrigation and Water Control. The principal objectives were to increase smallholder productive capacity, improve household food security and build the capacity of farmers to take over scheme operation and maintenance responsibilities by involving local communities. This was to be achieved through the development of a wide spectrum of simple low-cost technologies. The adoption of a participatory demand-driven approach was considered essential for ensuring sustainability. Under this approach, it was assumed that beneficiaries would be key partners involved in the selection of activities, their design and subsequent implementation.

In all, more than 80 000 were to benefit, including an appreciable number of women heads of households, and an area of 20 000 ha would be irrigated. The schemes included:

• Individually owned systems (7 percent of schemes). Small mobile pump schemes or various lift systems using seasonal river flow, stored water in ponds, or shallow aquifers.
• Community-based schemes with partial control of water (21 percent). Water harvesting, inland bottom valley schemes, and flood recession and control in West Africa.
• Communal gravity or pumping schemes (46 percent). Pumping systems operated by the villagers mostly developed in West Africa; gravity smallholder irrigation schemes based on the diversion of water stored in small dams (Ghana) and collective tube-wells with submersible pumping systems using deep groundwater (Senegal, Niger).

A review of the projects has produced some valuable insights into the challenges that this kind of intervention faces. Many of the projects were still being implemented and so a full assessment of their impact was premature. It was possible to see the progress made and identify the constraints that had hampered their implementation. Some of the key findings of the review included:

Impact on smallholder living conditions

In most cases the projects had a favourable impact in that they generally expanded the irrigated area, increased water availability and improved its use.

In regions where water supplies are dependable they enabled producers to use water for supplementary irrigation in the rainy season, to mitigate the effects of erratic rainfall on their traditional crops and, in the dry season, to produce high value cash crops, (Ghana, Ethiopia, Sudan and Niger). In Niger about 400 smallholders, who had difficulty in making the initial investment from their own resources, benefited significantly. Smallholders were able to buy small mobile pumps and in this way avoid a dependency on private wealthy pump owners, to whom they formerly had to pay one-third of their harvests in exchange for rental of similar equipment. In Senegal, farmers were able to increase their irrigated plots and the quality of construction was improved making maintenance easier. In Ghana, farmers benefited from irrigation during the dry season and saw this as the single most important factor in the achievement of household food security.

Impact on women's access to natural resources

Almost all projects had as a goal the introduction of increased equity and security in access to developed land especially by women. In Tanzania and Ghana, about a third of the irrigated plots were allocated to women. In addition, women reportedly contributed much of their labour towards scheme construction in order to assure themselves of a plot or extra income. However, the main actors who decided on land tenure issues were men rather than women (The Gambia, Mali, Senegal and Ghana). This still continues to be the case and the projects have not changed this.

Farmers' participation

The challenges of participation An engineer-agronomist told the anthropologist that he worked in close collaboration with the farmers every time he went out in the field. By this he actually meant that he was accompanied by the manager of the perimeter and the head of the county government. In another instance, at an introductory meeting of experts, who were going to work together on an irrigation project, the anthroplogist was introduced for the first time to the irrigation engineer who said, as they shook hands, `What are you doing here, we don't need you. We know everything we need to know about the farmers'. From Brown and Nooter (1992)

Conventional wisdom calls for farmers to become the principal partners and decision-makers in a project because this is perceived as the best way of ensuring ownership and development success. For this reason participatory approaches were a central plank of the planning phase. In practice, however, the projects were implemented under a project-led rather than a demand-driven approach. Participation was sacrificed in favour of speeding construction activities in order to develop as many schemes as possible. Even in cases where participatory processes were used, scheme design was usually in response to requests received by Project Managers, which had been made based on the perception landowners' and village chiefs' had of priorities, rather than those of the direct beneficiaries.

Generally, people implementing the projects did not have enough experience of participative techniques or were not convinced of their importance and paid only lip service to the process. Farmers were not always willing to contribute their labour for free, preferring instead to work on their farms or for off-farm income generating activities. This was contrary to the early expectations that people would be willing to invest their labour and this would engender a feeling of ownership. In many cases labour had to be paid for, which made translating theory into practice difficult. Time was also a major constraint. The demands of farmer participation, which require time to evolve, did not fit with the short time constraints that agencies normally place on project investment and construction possesses. The result was that the pace of development was much slower than originally anticipated.

Many projects failed to get farmers to participate. They were not convinced of the need to take full responsibility for bringing land into production, which in most cases they did not own. They would not take on the responsibility for operation and maintenance, which was viewed by farmers as a government provided source of employment and income rather than a self-help activity. Increasingly farmers are becoming aware that substantial support and subsidies can no longer be expected from the government and that they must begin to rely on their own efforts.

Suitable technologies

When dealing with communities living in harsh environments, existing production systems give farmers some basic security from starvation, even if they are also low yielding. Any external intervention should be developed around existing systems rather than bringing in new technologies. Expectations of cropping intensity yield and reduced water wastage need to be realistic. The outcome is determined more by the capacity of small-scale farmers to adopt new technologies, where services are erratic and market prices are high, rather than on what is potentially possible.

Although irrigation is meant to reduce risk, new technology introduces other risks that might expose farmers to a higher level of risk than before. Project reviews have shown that this extra risk falls exclusively on the farmers without being shared by the government. This has resulted in conflicts between farmers and credit institutions, and in farmers abandoning projects.

The most successful technologies proved to be those initiated to improve existing farming systems and which continue to be under the full control of the communities or individuals, such as: flood recession improvement, individual low-lift pumping systems, and abstraction and diversion of water from rivers. The risks increased considerably when larger scale pumping from tubewells for whole communities was introduced.

The idea of using `low-cost technology' is attractive to donors and farmers alike although it can and does create problems. There is usually an assumption that once a scheme has been planned, the design and construction is a straightforward process and the outcome is a well-engineered system of supply that farmers can use with confidence. This is not always the case and the phrase `low-cost technology' is in danger of becoming a euphemism for poor engineering design and construction. Canal embankments were often not compacted properly and so they did not last very long. Concrete was not prepared properly and compacted during placement with the result that it crumbled during use and so structures did not give adequate service. If the technology does not work properly, the seeds of failure are already sown.

The principal reason for much of the poor engineering was a serious shortage of suitably trained irrigation engineers and technicians who could undertake design work and supervise construction properly so as to produce sound engineering works.

Prospects for development

Many sub-Saharan Africa countries are becoming increasingly dynamic in response to more liberal marketing and pricing policies and greater competition between private sector interests. There are concerns that while the state may withdraw from a number of functions that may be better handled by the private sector, assistance for smallholder irrigation schemes targeting the poorest of the rural population can be expected to remain an important area for public sector investment and support. Without targeted assistance, at least during a transition period, the rural poor will become increasingly vulnerable to the new emphasis on an `open economy' environment and hopelessly marginalized.

Simple and appropriate water control technologies are becoming better known, and some decline in unit costs of development can be expected as smallholder irrigation systems gain greater credibility and farmers' awareness of their role and responsibilities in relation to water management increases. The institutional capacity of governments, NGOs and the private sector to work together is also improving. These factors would support continued investments in smallholder irrigation, especially since without investment in water infrastructure the prospects for increasing food production and improving food security are remote in many countries. Table 5 summarises the technical and institutional issues and gives an indication of costs.

Table 5: Summary of IFAD funded smallholder developments

	Traditional Community based schemes	Individuals and private schemes	Inter-village and communities schemes
Type and nature of systems	Water harvesting. Inland valley bottom. Flood recession	Small mobile petrol pumps. Various lift systems (treadle pumps, wind pumps, etc. Small collective pumps	Village irrigated. Diversion from rivers and small dams. Deep tubewells
Water users association	Not necessary/community matters	Not necessary, individuals or small groups responsibility	Necessary
Water management	Necessary because of impact on other communities	External monitoring necessary to avoid depletion of aquifers	Necessary for sustainability
Risks	Limited risk	Medium risk. No sharing of risk	High risk. Sharing mechanism needed
Average cost (US$/ha)	600 to 1 000	1 500 to 3 000	3 000 to 8 000

Summary of IFAD review

The following summarises some of the key lessons learned from the IFAD review:

• The identification of local community representatives or farmers' organizations capable of acting as responsible partners is a prerequisite for any smallholder irrigation development. Beneficiary involvement should be a partnership process where each partner has the right to agree or not; has the capacity to influence the decision and to share in the responsibility. In addition, attention should be given to matching the scale and scope of a project with the implementation capacity of local institutions and the farmers' capacity to participate.
• A detailed plan and procedures should be established so that implementers are well prepared for the tasks expected of them. This would not be a simple detailed list of physical activities, but rather a plan of who should do what, when and how.
• A distinction should be made between projects aiming to improve existing practices under the responsibility of traditional communities/villages from those introducing new ideas and technologies.
• More attention should be paid to site selection criteria, targeting and implementing procedures so that projects can proceed in a more realistic manner. Project design and implementation should be iterative. This is a social process and time is required to build on demand for assistance from groups of farmers and to create local support capacity sensitive to applying participative approaches.
• It is essential to have a thorough assessment of the labour that beneficiaries are willing to freely contribute to construction in relation to competing labour demands from existing farming systems and off-farm work. This input should not exceed 15 percent of the costs of work.
• Appropriate methods of construction appear to be those that optimise a mixed use of small local contractors and labour-intensive inputs provided by beneficiaries and include provision for independent professional checks on the design and construction standards to ensure these conform to specifications.
• Prepare an agreement between the sponsoring agency and the beneficiaries specifying their respective roles, contributions and obligations during construction and later for operation and maintenance. Cost recovery and financing mechanisms and procedures are still a major problem in many countries and still need to be worked out. Agree on appropriate mechanisms based on the technologies used, the cropping system and the organizational features of farmers.
• Forming Water Users' Associations (WUAs) and transferring project assets to them proved to be a long and difficult process incompatible with the short period of project implementation. This has administrative, financial and legal implications. Adequate support is needed for the creation of WUAs and for training members and office holders sufficiently early in the course of a project to ensure they are properly prepared to assume operation and maintenance responsibilities.
• Be prepared for the fact that effective beneficiary involvement and development of a sense of ownership is not always achievable for every type of scheme.
• Ensure that other support services (agricultural extension, input supplies, market access, rural finance, land allocation, etc.) are also in place. A realistic assessment of the institutional viability of securing these services during implementation is essential.

Existing situation

No recent and accurate information is available on the full extent of smallholder irrigation in sub-Saharan Africa. In this section an attempt is made, based on available data, to look at what has already been achieved and future possibilities on a country-by-country basis. The speed of past development may hold clues as to how fast we can realistically expect developments in the future.

The extent of irrigation development and how it has changed over the period 1960 to 1997 is summarised by Gleick (2000) on a country-by-country basis using FAO data sources. What is not readily available, however, is the current extent of smallholder irrigation relative to larger-scale developments. For this it is necessary to go back to 1986 when FAO estimated the split between the two. For this reason the FAO 1986 data are used to show the extent of smallholder irrigation in relation to the total developed in 1986. A comparison is made of the estimated potential for irrigation based on available land and water resources (Table 6). In addition to this, data from 1985 to 1997 (a 12 year period) quoted by Gleick have been used to calculate the average rate of irrigation development (smallholder plus large scale) in each country (see last column in Table 3). As this information provides an important insight into the capacity of each country to support irrigation development in terms of future rates of expansion, the countries have all been ranked according to their rate of development.

Table 6: Irrigated areas in sub-Saharan Africa in 1982

Source: FAO 1986 and Gleick 2000 ^1
1Countries are ranked in terms of their rate of irrigation development over the past 12 years.

The FAO consultation in 1986 concluded that in 32 of the 40 countries in sub-Saharan Africa, the first priority is for rainfed agricultural development. In eight countries Botswana, Burkina Faso, Kenya, Niger, Mali, Mauritania, Senegal and Somalia, irrigation is an essential part of any food security strategy because of the population pressures and the lack of rainfed capacity.

FAO data also show that sub-Saharan Africa includes 11 of the 16 nations of the world having less than 1 000 m³/head/year of water, a situation described as `absolute water scarcity' where food shortages are a constant threat and water shortage can only increase. (FAO 1995)

What is clear from the data is that the potential for further irrigation development in terms of land area and, to a lesser extent, water availability is very significant even allowing for the development that has taken place over the past 14 years (since 1986). Although there may be some differences over the definition of smallholder irrigation, the table clearly shows some 66 percent of irrigation development is in this sector. Since the emphasis in the intervening 14 years has also been on smallholder irrigation the true percentage may be even higher.

The average rates of development of irrigation over the past 12 years (1985-97) provide an interesting insight into what has been achieved in each country and what might be achieved in the future. Of the eight countries identified by FAO as having a dependency on irrigation the rate of development in Botswana, Mauritania and Senegal has been zero. Burkina Faso, Kenya, Niger, Mali and Somalia are in the top 14 best performing countries.

Unfortunately, even in the most irrigation active countries the rate of development is still very modest and over 50 percent have a rate less than 500 ha/year. In fact 13 countries show no growth at all. It may well be that irrigation is not as important in those countries. There are some on the list that would clearly benefit from some resurgence of interest in irrigation. The total estimated rate for the whole region is 43 600 ha/year - an average of 1 150 ha/year for each country.

Some countries appear to have encouraging rates of development (e.g. Tanzania, Nigeria, Niger, Zimbabwe and South Africa - the latter two are special as they have substantial private irrigation sectors). The question is - can this rate be sustained or increased in the future? Many development processes follow a classic `S' curve and irrigation is no exception. The curve usually has three distinct phases:

• A preliminary lag phase at the start of a new initiative when the ideas and technologies are relatively unknown and untried.
• An intermediate, rapid growth phase which takes place when the area under irrigation is being expanded very fast with active farmer participation using by now well tried and tested techniques.
• A tailing off into a flatter section when the available land or water is nearly used up and any further developments are on marginal areas with less potential benefit to the farmer.

Data from five countries over the period 1960 to 1997 demonstrate this, see Figure 4. Two countries are in the lower development rate group and three are in the higher group. In each case it is apparent that development has past through the rapid growth phase and is in the tailing off phase. These data are typical of all the countries listed in Table 6, which all exhibit a similar trend.

Figure 4: Rates of irrigation development in selected countries

The realities of these development curves add to the view that it is factors other than natural resources that constrain development. These include marketing, infrastructure, availability of fuel and spare parts, social background, labour availability, pricing policy, population density and land availability as well as the knowledge and expertise in irrigation technology.

Future potential

If the above data foretells the future then rates of development are likely to be slower than over the past 12 years unless the key constraints are identified in each country and steps taken to remove them.

If the current rate is sustained then an extra 1 million ha of irrigation could be brought into production in the next 25 years - only a 50 percent increase on what has already been achieved. To sustain this rate, let alone increase it, would require not only massive investment but also a considerable lead-time of several years to build the infrastructure and the absorptive capacity to support it. In view of this, perhaps a more realistic estimate might be half this - say 0.5 million ha.

Rates of development In general, areas with small reserves of land and high populations are likely to show the fastest change as farmers will be under pressure to intensify their agricultural production in the face of falling land availability. In Burundi the first record of the use of valley swampland appeared in 1936. By 1959 this had risen to 200 ha under dry season cultivation and by 1975 this had increased to 4 000 ha. By 1979 a survey recorded 5 700 ha of swamp irrigation, an increase of nearly 43 percent in four years (Gerard 1982 in Hekstra 1983). A similar growth curve was observed in Nigeria with an increase in traditional fadama irrigation from 120 000 ha in 1958 to 800 000 ha in 1978. This has gone on with little or no support from the Government. This growth in smallholder irrigation contrasts strongly with the actual development of formal large-scale irrigation in Nigeria. The lag phase has lasted more than 30 years. This is contrary to the predictions of both the Federal Department of Agriculture and the World Bank, that predicted areas of 320 000 ha and 125 000 ha respectively under irrigation by 1982. By contrast the actual total area under formal irrigation in 1982 was only 30 000 ha. The declared total irrigation in Nigeria in 1995 was 2.3 million ha and the majority of this is likely to be in the smallholder sector. Carter et al (1983).

Strong recommendations have been made by many leading experts that the process of development will inevitably be a slowly incremental requiring low investment sustained over a long period. Some argue that low world grain prices are the main cause of irrigation stagnation, put simply, irrigation does not pay well at the moment. Increases in grain prices would stimulate the markets and encourage farmers to produce more. The message is clear; it is not the availability of physical resources that is likely to constrain the development of irrigation in Africa but economic, social and institutional constraints that could hamper progress. Unless radical steps are taken to ease these constraints the result will be only a modest reduction of Africa's food problems over the next 25 years.