Previous Page Table of Contents Next Page

3. Why agricultural water productivity is important for the global water challenge

Productivity is a ratio between a unit of output and a unit of input. Here, the term water productivity is used exclusively to denote the amount or value of product over volume or value of water depleted or diverted. The value of the product might be expressed in different terms (biomass, grain, money). For example, the so-called ‘crop per drop’ approach focuses on the amount of product per unit of water. Another approach considers differences in the nutritional values of different crops, or that the same quantity of one crop feeds more people than the same quantity of another crop. When speaking of food security, it is important to account for such criteria (Renault and Wallender, 2000). Another concern is how to express the social benefit of agricultural water productivity. All the options that have been suggested can be summarized by the phrases ‘nutrient per drop’, ‘capita per drop’, ‘jobs per drop’, and ‘sustainable livelihoods per drop’. There is no unique definition of productivity and the value considered for the numerator might depend on the focus as well as the availability of data. However, water productivity defined as kilogram per drop is a useful concept when comparing the productivity of water in different parts of the same system or river basin and also when comparing the productivity of water in agriculture with other possible uses of water.

Crop water production is governed only by transpiration. As it is difficult to separate transpiration from evaporation from the soil surface between the plants (which does not contribute directly to crop production), defining crop water productivity using evapotranspiration rather than transpiration makes practical sense at field and system level. In irrigated agriculture in saline areas, the leaching requirement, i.e. the amount of water that needs to percolate to maintain rootzone salinity at a satisfactory level, should also be included together with evapotranspiration in the amount of water that is necessarily depleted during plant growth. Other non-productive but beneficial uses could be included. Examples are evapotranspiration by windbreaks, cover crops, and the water used in wetting seedbeds to enhance germination.

The question of considering water losses from seepage and field percolation as consumption does not receive a unique response. If this water is of no use downstream or if it generates further pollution such as that resulting from geological salt leaching (e.g. San Joaquin Valley, California, the United States of America), then it must be accounted for as consumption. Solutions to minimize these losses, such as canal lining or water improvement application, then have a positive effect on productivity. However, from a broader environmental point of view, it can be important to consider the impact of the outflow of an irrigation system on the overall productivity of an ecosystem.

As with the numerator, the choice of the denominator (which drops to be included) should depend on the scale, the point of view and the focus. At basin level, the choice might be between water diverted from the source and the same minus water restored, whereas at field level one might consider useful rain, irrigation water and supplemental irrigation.

Spatial variability of water productivity

Reported data on water productivity with respect to evapotranspiration (WPET) show considerable variation, e.g. wheat 0.6-1.9 kg/m3, maize 1.2-2.3 kg/m3, rice 0.5-1.1 kg/m3, forage sorghum 7-8 kg/m3 and potato tubers 6.2-11.6 kg/m3, with incidental outliers obtained under experimental conditions. Data on field-level water productivity per unit of water applied (WPirrig), as reported in the literature, are lower than WPET and vary over an even wider range. For example, grain WPirrig for rice varied from 0.05 to 0.6 kg/m3, for sorghum from 0.05 to 0.3 kg/m3 and for maize from 0.2 to 0.8 kg/m3. The variability occurs because data were collected in different environments and under different crop management conditions. These affected the yield and the amount of water supplied (Kijne et al., forthcoming). Furthermore, it is often difficult to determine the real crop yield over a large area, e.g. the size of a large irrigation system. When asked for yield figures, individual farmers are likely to give a figure that depends on the situation. For a loan application, they may overstate the yield, whereas for payment of a debt or a tariff, they will probably understate the yield obtained. Vegetable yields of vegetables may change every day, and unless good records are kept, no one will know exactly how much was harvested during the total harvest period. Yields expressed in monetary terms are more doubtful as prices on the local market may fluctuate considerably over time (FAO, 2002d).

Plate 7 A farmer works on an irrigation canal (Mexico)


Nevertheless, water productivity data across scales are useful in assessing whether water drained from upstream is reused effectively downstream. However, there are few reliable data on water productivity at different scale levels within the same system. A study using remote sensing and GIS technologies assessed crop WPET at various irrigation system scales in the Indus Basin in Pakistan (Bastiaanssen et al., 2003). Crop water productivity was found to vary significantly at the scale of small canal command areas. When water productivity was aggregated for canal command areas, the highest water productivity values decreased gradually. Their variability also decreased until at a scale of about 6 million ha water productivity tended to a low value of about 0.6 kg/m3. This arose because at the larger scale, canal commands with less fertile or saline soils and with less canal water and poorer quality groundwater were included in the average.

Box 4 presents data illustrating the productivity of water in economic terms.

Box 4 Water productivity in economic terms

Source: Merrett, 1997; Molden et al., 2001

Data are available for agricultural water productivity in economic terms for Jordan. Water productivity ranged from US$0.3/m3 for potato to US$0.03/m3 for wheat. The average value for agricultural products was US$0.19/m3 and for industrial products US$7.5/m3. The IWMI analysed economic water productivity data from two irrigation systems in South Asia. The values for wheat production ranged from US$0.07 to 0.17/m3. Average systemwide water productivity values of US$0.10 and 0.15/m3 were reported for two other systems in South Asia. Systemwide values for a total of 23 irrigation systems in 11 countries in Asia, Africa and Latin America ranged from US$0.03/m3 (for a system in India) to US$0.91/m3 (for one in Burkina Faso), with an overall average of US$0.25/m3. Comparison with the most recent cost of about US$0.50/m3 for desalinated seawater illustrates that this source of water is too expensive for virtually all agricultural production. However, its cost has come down to about one-tenth of what it was 20 years ago. Further improvements in the technology of seawater desalination are likely. Its cost is also likely to continue falling provided that as energy remains cheap.

The substantial increase of water productivity in agriculture

Despite concerns about the technical inefficiency of water use in agriculture, water productivity increased by at least 100 percent between 1961 and 2001. The major factor behind this growth has been yield increase. For many crops, the yield increase has occurred without increased water consumption, and sometimes with even less water given the increase in the harvesting index. Example of crops for which water consumption experienced little if any variation during these years are rice (mostly irrigated) and wheat (mostly rainfed), for which the recorded increases worldwide amount to 100 and 160 percent respectively. At the global level, the increase in water consumption for agriculture in the past 40 years has been 800 km3 (Shiklomanov, 2000) while world population has doubled to 6 000 million. Considering that the arable rainfed area has not increased, one can conclude that with an additional 800 km3 of water the world has been able to feed an additional 3 000 million people. This gives a rough estimate of 0.720 m3/d/capita. This figure is low compared to the estimated global average for 2000 of 2.4 m3/d/capita, which includes water for food at field level not including water losses. This is a good indicator of the significant productivity gain recorded in agriculture; a gain that has enabled the world to accommodate the doubling of the population and also increase intake.

As a whole, one can estimate that the water needs for food per capita halved between 1961 and 2001 from about 6 m3/d to less than 3 m3/d (Renault, 2003).

The importance of water needs for food makes any small relative gain in this sector equivalent to a significant gain for other uses. For example, given the water needs for capita in 2000, a 1-percent increase in water productivity in food production generates a potential of water use of 24 litres/d/capita. In order to produce the equivalent of the domestic water supply, a gain of 10 percent in agricultural water productivity would be required, which is a matter of years. Therefore, it can be argued that investing in agriculture and in agricultural water is the best avenue for freeing water for other purposes.

However, future agricultural gains will need to be split into several components: (i) compensation for the reduction of agricultural production areas as a result of urban encroachment, soil degradation, and the depletion of water resource availability or access (groundwater); (ii) increased water access for the rural poor and vulnerable groups; (iii) generation of wealthier farming systems; and (iv) freezing water for other uses including the environment.

Key principles for improving water productivity

The key principles for improving water productivity at field, farm and basin level, which apply regardless of whether the crop is grown under rainfed or irrigated conditions, are: (i) increase the marketable yield of the crop for each unit of water transpired by it; (ii) reduce all outflows (e.g. drainage, seepage and percolation), including evaporative outflows other than the crop stomatal transpiration; and (iii) increase the effective use of rainfall, stored water, and water of marginal quality.

The first principle relates to the need to increase crop yields or values. The second one aims to decrease all ‘losses’ except crop transpiration. Its phrasing does not imply that it will be impossible to increase water productivity by reducing stomatal transpiration. It is conceivable that plant breeding may find ways to overcome this constraint. The third principle aims at making use of alternative water resources. The second and third principles should be considered parts of basinwide integrated water resource management (IWRM) for water productivity improvement. IWRM recognizes the essential role of institutions and policies in ensuring that upstream interventions are not made at the expense of downstream water users.

Plate 8 Food transfer can be considered equivalent to transfer of “virtual water” (Somalia)

FAO/20430/A. PROTO

These three principles apply at all scales, from plant to field and agro-ecological levels. However, options and practices associated with these principles require different approaches and technologies at different spatial scales.

Enhancing water productivity at plant level

Plant-level options rely mainly on germplasm improvements, e.g. improving seedling vigour, increasing rooting depth, increasing the harvest index (the marketable part of the plant as part of its total biomass), and enhancing photosynthetic efficiency. The most significant improvements in yield stability have usually resulted from breeding programmes to develop an appropriate growing cycle such that the duration of the vegetative and reproductive periods are well matched with the expected water supply or with the absence of crop hazards. Planting, flowering and maturation dates are important in matching the period of maximum crop growth with the time when the saturation vapour pressure deficit is low. The periods of maximum crop growth may be optimized by means of breeding technology. Improved varieties with a deeper rooting system contribute to drought avoidance and the effective use of water stored in the soil profile. Drought escape and increasing drought tolerance are also important strategies for increasing water productivity (Box 5). Daylength-insensitive varieties of short to medium duration (90-120 d) enabled crops, such as wheat, rice and maize varieties developed as part of the green revolution, to increase water productivity by escaping late-season drought that adversely affects flowering and grain development. The modern rice varieties have about a threefold increase in water productivity compared with traditional varieties (Tuong, 1999). Progress in extending these achievements to other crops has been considerable and will probably accelerate following the recent identification of the underlying genes (Bennett, 2003). Genetic engineering, if properly integrated in breeding programmes and applied in a safe manner, can further contribute to the development of drought tolerant varieties and to increasing the water use efficiency.

Box 5 Real impacts of virtual water on water savings

Source: Renault, 2003; Zimmer and Renault, 2003

Exchanges of virtual water through food trade first captured the attention of experts in the Near East, where water is scarce (Allan, 1999) and imports represent considerable water savings. The value of virtual water of a food product is the inverse of water productivity. It is defined as the amount of water per unit of food that is or would be consumed during its production process.

Virtual water trade generates water savings for importing countries. It also generates global real water savings because of the differential in water productivity between the producing and the exporting countries. For example, transporting 1 kg of maize from France (taken as representative of maize exporting countries for water productivity) to Egypt transforms an amount of water of about 0.6 m3 into 1.12 m3, which represents globally a real water saving of 0.52 m3 per kilogram traded. In 2000, the maize imports in Egypt and the related virtual water transfer thus generated a global water saving of about 2 700 million m3. The global real water saving is significant: a first estimate shows that water savings from virtual water transfer through food trade amounts to 385 000 million m3 (Oki et al., 2003).

Food storage also generates real water savings. For example, in the Syrian Arab Republic, 1988 was a good year for the cereal production with yields of 1.6 tonnes/ha, leading to a surplus. Thus, 1.9 million tonnes of cereals were stored during that year. The following year was very dry, and the cereal yield dropped to 0.4 tonnes/ha. About 1.2 million tonnes of cereals were then withdrawn from storage to complement internal production and imports. Based on the water productivities recorded for these years (Oweis, 1997), the estimated value of virtual water was to 1 and 3.33 m3/kg respectively. Therefore, the use of 1.2 million tonnes of cereals from storage in 1989 is equivalent to 4 000 million m3 of virtual water. For the two-year period of reference (1988-89), some 2 800 million m3 of water was saved by the food storage capacity.

Globally, the trade in virtual water is rising rapidly. It increased in absolute value from about 450 km3 in 1961 to 1 340 km3 in 2000, reaching 26 percent of the total water required for food including equivalence for sea products and sea fish. This value is shared evenly between energy, fat and protein products.

Raising water productivity at field level

Improved practices at field level relate to changes in crop, soil and water management. They include: selecting appropriate crops and cultivars; planting methods (e.g. on raised beds); minimum tillage; timely irrigation to synchronize water application with the most sensitive growing periods; nutrient management; drip irrigation; and improved drainage for water table control.

Water depletion occurs when water evaporates from moist soil, from puddles between rows and before crop establishment. All cultural and agronomic practices that reduce these losses, such as different row spacings and the application of mulches, improve water productivity. The irrigation method also affects these evaporative losses. Drip irrigation causes much less soil wetting than sprinkler irrigation. The significance of soil improvement in enhancing water productivity is often ignored. However, integrated crop and resource management practices, such as improved nutrient management, can increase water productivity by raising the yield proportionally more than it increases evapotranspiration. This principle applies to both irrigated and rainfed agriculture. Integrated weed and integrated pest management have also contributed effectively to yield increases.

Plate 9 Model of integrated fish farm. Combination of fish ponds with ducks (Lao People’s Democratic Republic)

FAO/20906/K. PRATT

One of the field-level methods for increasing water productivity is deficit irrigation, where deliberately less water is applied than that required to meet the full crop water demand. The prescribed water deficit should result in a small yield reduction that is less than the concomitant reduction in transpiration. Therefore, it causes a gain in water productivity per unit of water transpired. In addition, it could lower production costs if one or more irrigations could be eliminated. For deficit irrigation to be successful, farmers need to know the deficit that can be allowed at each of the growth stages and the level of water stress that already exists in the rootzone. Most importantly, they need to have control over the timing and amount of irrigations. Deficit irrigation carries considerable risk for the farmers where water supplies are uncertain, as is the case with rainfall or unreliable irrigation supplies. Where water availability falls below a certain level, the value of the crop can fall to zero, either because the crop dies or because the product is of such low quality as to be unmarketable. When water is scarce, farmers could reduce the irrigation as appropriate to maximize returns to water if they have control over the timing and amount of irrigations.

This degree of flexibility is usually the case with sprinkler and drip irrigation, and also with pumped groundwater if the farmer owns the pump. A totally flexible delivery system for surface irrigation in large irrigation systems is expensive because of the required overcapacity in the conveyance system.

The trade-off between reduced yield and higher water productivity needs to be quantified in economic terms before recommending deficit irrigation (and other water-saving irrigations in rice production).

The often cited low water productivity per unit of water supply in rice cultivation derives from considering as losses the percolation resulting from the standing water layer on the field surface. However, this water is often recycled, and rice water productivity generally compares well with that of a dry cereal. Nevertheless, water-saving irrigation techniques such as saturated soil culture and alternate wetting and drying can reduce the unproductive water outflows drastically and increase water productivity. These techniques generally lead to some yield decline in the current lowland rice high-yielding varieties (Box 6). However, some experiments are reporting substantial yield increases for local varieties (Deichert and Saing Koma, 2002) using a technique called system rice intensification (SRI), a technique which originated in Madagascar (de Laulanié H., 1992). Here again there is no unique response; the fit with local resources and capacity is the most important feature to account for. Without anticipating results of current investigations in many countries, it seems that the potential of the SRI technique for the poor to increase the productivity of scarce land and water is significant provided that enough family labour is available. Other approaches are being researched as part of efforts to increase water productivity without sacrificing yield. One of these is to develop so-called aerobic rice systems that allow rice cultivation in non-flooded conditions. The development of these new rice varieties is essential if rice is to be grown like other irrigated upland crops and the deep percolation associated with paddy rice is to be avoided.

Box 6 Water-saving irrigation technologies in rice production

Source: IRRI, 2002

Exploring ways of producing more rice with less water is essential for food security in Asia while also protecting the environment. The International Rice Research Institute (IRRI) has studied various field-level water-saving technologies, e.g. alternate wetting and drying; SRI; saturated soil culture; aerobic rice; and ground-cover systems. Each of these techniques reduces one or more of the unproductive water outflows (e.g. seepage, percolation, and evaporation) and hence increases water productivity. However, they also introduce periods in which the soil is not flooded or not even saturated, which usually leads to yield decline. Recent results from northern China and the Philippines indicate that with current germplasm and management technologies, aerobic rice yields are about 40 percent lower and reduce water requirements by about 60 percent compared with flooded lowland systems.

The shift from flooded systems to partly aerobic (non-saturated) conditions also has profound effects on soil organic matter turnover, nutrient dynamics, carbon sequestration, weed ecology and greenhouse gas emissions. Whereas some of these changes are positive, others, such as the release of nitrous oxide and the decline in organic matter, are perceived as negative effects. The challenge is to balance the negative and positive effects through the development of effective integrated water-saving technologies that can ensure the sustainability of rice-based ecosystems and environmental services.

Water-related problems in rainfed agriculture are often related to large spatial and temporal rainfall variability rather than low cumulative volumes of rainfall. The overall result of rainfall unpredictability is a high risk for meteorological droughts and intraseasonal dry spells (Rockström et al., 2003). Bridging crop water deficits during dry spells through supplementary irrigation stabilizes production and increases both production and water productivity dramatically if water is applied at the moisture-sensitive stages of plant growth.

Water harvesting for agriculture involves a storage reservoir, while in runoff farming the collected runoff is applied directly to the cultivated area. Either way, the investments in the construction of the ditches that take the runoff to the storage reservoir and of the reservoir itself are relatively small. Maintaining these structures may be more difficult if heavy rains periodically wash them away. Many factors affect the success of rainwater harvesting. These include: the method used for runoff collection and storage; the topography; the soil characteristics (especially the infiltration rate); the choice of crop to be planted; fertilizer availability; and the effectiveness of the soil crust in the catchment area. However, probably more important than any of these physical parameters is the involvement of the beneficiaries in the design and implementation of the water harvesting structures (Box 7).

Box 7 A soil and water conservation project in Burkina Faso

Source: Oweiss et al., 1999

Until the early 1980s, most soil and water conservation projects in Burkina Faso had failed dramatically. From 1962-65, heavy machinery was used to treat entire catchments in the Yatenga Region of the country’s Central Plateau with earthen bunds. Although the project, which treated 120 000 ha in 2.5 dry seasons, was well-conceived technically, the land users were not involved, and they were not at all interested in what had been constructed. From 1972-1986, several donor agencies funded a soil and water conservation project based on a more participatory approach. However, once again, the land users were not willing to maintain the earth bunds because of the high maintenance requirement, lack of benefits and other reasons. As a result, most of the bunds had disappeared entirely after 3-5 years.

An NGO-supported agroforestry project (1979-1981) in the Yatenga Region tested a number of simple soil and water-conservation/water-harvesting techniques and asked the villagers to evaluate the techniques. They expressed a preference for contour stone bunds. The project also initiated training programmes at village level that taught farmers how to use a water tube level, so enabling them to determine contour lines more accurately. In the Yatenga and other parts of the Central Plateau, tens of thousands of hectares have now been treated with contour stone bunds.

The main reason why the farmers adopted contour stone bunds and improved traditional planting pits (a technique developed by a local farmer in which water and fertilizers are mixed together) is that they produced immediate and substantial yield increases. On land that is already cultivated, the construction of the contour stone bunds is estimated to increase yields by 40 percent.

Socio-economic assessments of water harvesting and supplementary irrigation are rare. It is recognized that sustainable increases in water productivity by water harvesting can only be achieved through a combination of farmer training, water conservation, supplementary irrigation, better crop selection, improved agronomic practices, and political and institutional interventions. Planning (and economic assessment) should consider explicitly the short-term effect and longer-term implications of hydrological changes brought about by water harvesting on downstream water users.

Plate 10 Members of the village committee of Ankofafa protecting a maize field (Madagascar)


This paper has mentioned a number of practices that have the potential to enhance water productivity/The question now is one of how to stimulate the adoption of these techniques and their adaptation for local conditions. The importance of farmer participation and empowerment through the organization of WUAs in irrigation management is well accepted and they are widely established. However, less well known is the feasibility and advantage of using the same form of farmers’ associations for the purpose of introducing collectively improved cultural practices, such as minimum tillage or raised beds. Adoption of a range of water-productivity-enhancing practices by a large number of individuals should be stimulated through community-level interventions in order to ensure that opportunities to divert unallocated water to other productive uses are not missed.

Accounting for water productivity at system and basin level

Changing the focus from the field level to system and river-basin level changes the relative importance of the various water management processes. At the larger scale, the effect of agriculture on other water users, human health and the environment becomes at least as important as production issues.

Options for improving water productivity at the agro-ecological or river-basin level are found in: better land-use planning; better use of medium-term weather forecasts; improved irrigation scheduling to account for rainfall variability; and conjunctive management of various sources of water, including water of poorer quality where appropriate. Therefore, integrating germplasm improvement and resource management is crucial in the enhancement of water productivity at the field scale and above.

Gains in water productivity are possible by providing more reliable irrigation supplies, e.g. through precision technology and the introduction of on-demand delivery of irrigation supplies (Chapter 6). However, an increase in water productivity may or may not result in greater economic or social benefits. The social benefits represent the benefits to society resulting from the water-productivity-enhancing interventions. Water in the rural areas of developing countries has many uses. Thus, water is both a public and a social good, a fact that complicates value calculations. These many uses of water include: the production of timber, firewood and fibre; and raising fish and livestock. Non-agricultural uses of water include domestic (drinking and bathing) and environmental uses.

An IWMI study of an irrigation system in Kirindi Oya in southern Sri Lanka illustrates the importance of the multiple roles of water in agriculture (Renault et al., 2000). The study found that at system level crops consumed only 23 percent of the total water supply, including both rainfall and external irrigation water. Of the remainder, 8 percent was used for grazing land, 6 percent evaporated from the reservoir, 16 percent was lost to the sea, 3 percent drained into lagoons, while as much as 44 percent of the water supply went to perennial vegetation that had developed since the construction of the scheme. This perennial vegetation was there because of irrigation seepage and recharge of the shallow groundwater. Tree growth is important to the people living in the area as it provides them with shade and thus improves their environment. In this project, as well as in many places in southern India, it also provides income from coconut and materials for construction (beams and ropes). Other trees are important for additional nutritional values (fruits) and some are crucial for their medicinal properties. A changeover to total control of irrigation outflow in order to increase water productivity would cause the collapse of the entire local agroforestry system (FAO, 2002d).

Another example of the economic and social benefits of agroforestry is a project located along the Niger River in Mali. In this project, trees were planted on the bunds of rice fields, and also in the middle of the rice fields without affecting rice yields adversely. In this remote arid part of Mali, the value of the wooden poles of seven-year-old eucalyptus trees was so high that the farmers could pay for the O&M of the irrigation system from the sale of the trees. In another irrigation system, in southwest Burkina Faso, oil-palm and fruit trees were combined successfully with irrigated crops (mainly maize, groundnuts and industrial tomatoes). Trees were planted on ridges or on the boundaries between parcels. On the sandy, percolating soils of the irrigation system, the trees produced an important amount of complementary food and income, while the impact on the main crop was minimal (FAO, 2002d). Box 8 presents a case where traditional agriculture had greater benefits for society than did large-scale irrigation.

Box 8 Benefits from traditional floodplain agriculture compared with large-scale irrigated agriculture

Source: IUCN, 2000

The estimated value of the Hadejia-Jama’ara floodplain use in northern Nigeria indicates that traditional practices provide higher benefits than crops grown on the Kano irrigation project. Benefits derived from firewood, recession agriculture, fishing and pastoralism were estimated at US$12/litre of water, compared with US$0.04/litre for benefits from the irrigation project. This evaluation is important for the region as more than half of the floodplains have been lost to drought and upstream reservoirs.

Even without accounting for such services as wildlife habitat, the wetland is more valuable to more people in its current state than after conversion to large-scale irrigated agriculture.

These examples point out that not all measures to increase water productivity are appropriate in all circumstances. It is essential to consider the various uses of water in agriculture before measures are introduced that would increase water productivity at the expense of other benefits from the same source of water, especially those benefits that accrue to the local poor and landless people.

Policy tools for promoting water productivity gains

Using price policies to promote the economic productivity of water requires significant government intervention in order to ensure that equity of access to water and public-good issues are covered adequately (Barker et al., 2003; Rogers et al., 2002). Some studies in the Indian subcontinent and elsewhere have suggested that the price for water that would be required to affect demand substantially would be about ten times the charge required to cover the O&M of the irrigation system. A charge sufficient to cover O&M would have a minimal effect on water demand. Moreover, introducing volumetric charges for irrigation water is difficult and involves considerable expense for the installation of measuring structures and for fraud prevention (Perry, 2001). Last, in most rice-based systems in Asia, volumetric charging at individual user level or even group level is unsuitable given the permanent overflow and recycling water flows throughout the command area.

The groundwater market in India illustrates the perhaps unintended impact of government policies on the availability of water to farmers and others. Farmers in Gujarat paid about four times as much for pumped groundwater compared with farmers in Punjab and Uttar Pradesh. This difference was attributed to: (i) differences in the way farmers were charged for the electric power to run their pumps (flat rate versus per unit consumed); (ii) the tubewell spacing policy in Gujarat that gave each tubewell owner a monopoly over some 203 ha; and (iii) the scarcity of public tubewells in Gujarat, which also reduced competition among groundwater suppliers. The high prices for tubewell water in Gujarat discriminated against small and poor farmers. However, some simple changes in water policies for power pricing, tubewell spacing and public tubewells could transform groundwater markets in Gujarat into powerful instruments for small-farmer development (Shah, 1985).

Aiming for the highest economic productivity of water in agriculture may conflict with the political desire for national food security. More often than not, the economic productivity of water in growing staple crops is less than that for growing vegetables or flowers for export markets. Crop substitution involves switching high water-consuming crops for less water-consuming crops or for crops with higher economic productivity. The approach provides a strategy for increasing crop water productivity at the agro-ecological system level as well as at the global level (Box 5).

Policies and incentives are important in the adoption of changes from traditional agronomic and cultural practices (FAO, 2001a). However, it is necessary to identify the types of policies and incentives that will work best. Experience with conservation agriculture indicates that the short-term interests of the farmers often differ from the long-term interests of society and that the financial benefits that accrue from changes in cultural practices often take a long time to materialize. In addition, although there are large differences between individual farms, external factors also play a role, e.g. the transmission of information (via policy-related activities and social processes). Of particular importance is the fact that the inconsistent and sometimes contradictory results from studies on the adoption of new practices suggests that the decision-making process is highly variable. This decision-making process needs to be understood more fully as it will affect the lead time from study to field practice. This lead time is often unacceptably long considering the urgent character of water-scarcity problems. Experience from participatory research and extension could help reduce this lead time.

Previous Page Top of Page Next Page