Introductory papers

Overview of large irrigation systems in Southeast Asia
Zhijun Chen¹

Abstract

Most of the large irrigation systems in Southeast Asia have been designed for rice irrigation under supply-driven mode. Despite their huge contribution to agriculture production and local social and economic development, there has been a general consensus that these large rice irrigation systems have not lived up to expectations because of a legacy of poor institutional arrangements and system design, degraded infrastructure, poor management and stagnation in the face of rapid transformations of agriculture and pressure on their water supply. Population growth, economic development, urbanization and globalization have been and will continue to be the main drivers of changes in the agriculture and water sectors. Southeast Asia is expected to be the subregion where the greatest urban growth will occur over the next thirty years. Agriculture water use is facing more competition. In the meantime, diversification, commercialization and modernization of agriculture call for more flexible, high-quality, user-oriented water services. Large rice irrigation systems are facing multiple challenges.

During the past decades, responses have been made and initiatives have been taken by the international community, government departments and farmers on field water management improvement, institutional reform and governance strengthening, systems rehabilitation and irrigation modernization. So far the progress has been modest and, in general, the performance of large rice irrigation systems in Southeast Asia continues to be low in terms of control, water productivity, yields, and quality of service delivery to farmers. The differences between stated policies and actual practices are generally large. Significant underinvestment in operation and maintenance and poor management continue to be the norm rather than the exception. It is therefore necessary to review the experiences gained and the lessons learned, to analyze current strategies and programmes, and to propose further options.

This paper, first, reviews the historical evolution of irrigation and the current status and constraints of large rice irrigation systems in Southeast Asia; second, analyzes the changing trends of the agriculture and water sectors in Southeast Asia and their implications for large rice irrigation systems; third, discusses the responses and initiatives (both past and present) of government departments and the international community; fourth, evaluates the outcomes and constraints; and fifth, identifies issues to be addressed and options to be taken.

It is concluded that to meet the future challenges better, multiple options and systematic approaches are needed to transform large rice irrigation systems in Southeast Asia from supply-driven to demand-driven responsive systems, to improve their water service in terms of reliability, equity, flexibility and multiple use, and to enable farmers to boost agricultural and water productivity and be more responsive to market opportunities.

1. Large irrigation systems in Southeast Asia

Historical evolution

The development of large irrigation systems has a long history in Asia. The world-famous Dujiangyan irrigation system was built 2 000 years ago in present-day Sichuan Province of China, and it is still functioning with a current irrigated area of 670 000 ha. Pyu, located in present-day Myanmar during the sixth to eighth centuries AD, was one of the first kingdoms in Southeast Asia that established a complex irrigation system. Large-scale, state-centred antecedent irrigation systems in agrarian societies enabled food surpluses and released labour for other cultural activities and were associated with the formation of many of the early powerful kingdoms. Barker and Molle (2004) in their review of the development of modern Asian irrigation distinguished three time periods in terms of their geopolitical significance: the Colonial Era (1850 to 1945), the Cold War Era (1946 to 1989) and the New Era of Globalization (1990 onward).

The Colonial Era was characterized by irrigation expansion by the colonial powers, mainly through river diversion, to control famine, unrest, or revolt and to extract surplus, such as the large rice expansion in Indonesia by the Dutch from 1900 to 1940, expansion of canals and paddy fields in Viet Nam from the 1860s to the 1930s, and the reclamation of the Irrawaddy delta in Myanmar. There are some exceptions: the Chao Phraya Delta in Thailand was reclaimed between 1850 and the mid-twentieth century in the absence of formal colonization; many irrigation advances in China that occurred centuries ago were not influenced by colonial programmes.

The Cold War Era was characterized by rapid irrigation expansion mainly through construction of dams, reservoirs and canal systems, and later on followed by public and privately financed groundwater exploitation. From 1950 to the late 1970s, 50 percent or more of the agricultural budget in many Asian countries was devoted to irrigation. From the 1970s to the 1980s, more than half of the World Bank spending for agriculture was for irrigation. In the late 1970s and early 1980s, lending for irrigation by the World Bank and ADB reached a peak of over one billion US dollars per year in constant 1980 dollars, but fell to less than half that level by the late 1980s because of the sharp drop of cereal grain prices in the mid-1980s, the rise of construction costs and the growing opposition of environmentalists.

The New Era of Globalization is characterized by the control and management of water for agricultural and non-agricultural uses. As water has become scarce as a result of growing demand, water pollution has been increasing, and yield gains in cereal production in most of Asia have slowed or leveled off, profits from cereal grain prices have declined and Asian farm households rely increasingly on income from non-farm sources. The period of the rapid expansion of the irrigated area through either construction of surface irrigation systems or exploitation of groundwater has seemingly come to an end. Attention has turned to improving the management and performance of existing irrigation systems both to reduce the financial burden and to allow an increasing share of water to be diverted to non-agricultural uses.

Table 1. Evolution of publicly managed irrigation in South and Southeast Asia

Source: Barker & Molle (2004).

The evolution of irrigation in Southeast Asia generally followed the regional path, but with varied paces and some diversification. The construction phase had reached a peak by the end of the 1970s in Indonesia, Malaysia, the Philippines, Thailand and China. After 1985, there was a significant decline in the rate of growth in these countries. But in recent years, irrigation investment has made a comeback in Malaysia, China and Thailand, mainly focused on large system rehabilitation and new small system development. In Cambodia, Lao PDR, Myanmar and Viet Nam there has been continued strong growth during the past two decades and various irrigation systems have been under design and construction.

Table 2. Annual growth in irrigated area in Asia and in the countries of its subregions, 1961-1999

Notes: Calculations are based on three-year averages centred on the years shown.
SEA I = Indonesia, Malaysia, the Philippines and Thailand
SEA II = Cambodia, Lao PDR, Myanmar and Viet Nam
Source: Barker & Molle (2004).

Recent developments

During the past decade, irrigation evolution in Southeast Asia has shown both a general trend and diversification. The general trend is that software systems, including water management, system management, legislation, institutions, and participation have been highlighted in all countries. The diversification is that different group of countries have taken different actions on hardware systems corresponding to their specific resources conditions and agriculture and economic development requirements. In Indonesia, Philippines and Thailand the major focus has been the innovation in software systems of existing irrigation schemes for better exploiting their potential, with some small-scale irrigation system development. Indonesia has conducted a large-scale irrigation rehabilitation and development programme since the 1960s. Since that time, the government has focused on effective and efficient irrigation water management through increased reliance on institutional strengthening and effective interagency coordination. The Philippines has been continuously struggling to improve water availability, allocation, regulation and irrigation performance through irrigation sector reform and policy innovation such as “no payment no irrigation” (Domingo, 2005). Thailand has been driven by the recent increasing rice prices and by droughts to exploit better the potential of existing large irrigation schemes by adopting participatory irrigation management (PIM), changing irrigation schedules, combined with groundwater exploitation.

In Cambodia, China and Malaysia there has been a comeback of large irrigation investments, and the major focus has been on modernization of existing large irrigation systems through joint efforts addressing both software and hardware issues. In 2000, Cambodia adopted participatory irrigation management and development policy (PIMD), and has been focusing on its implementation since then. The policy requests irrigation management transfer from government to water users associations. A prerequisite for the successful transfer is proper physical upgrading before the transfer. In support of this policy, thirteen donors, mostly international funding agencies, including the Asian Development Bank (ADB), the European Union (EU) and the World Bank have conducted 20 projects on irrigation infrastructure rehabilitation, followed by management transfer and WUA empowerment.

Table 3. Irrigation in Indonesia in 1966 and 1989

Source: Country paper from Indonesia (in this publication).

The Chinese Government tackles water saving as “a revolutionary option”. While conducting nationwide water legislation innovation, institutional reform and technological dissemination, the nationwide “water-saving rehabilitation of large-scale irrigation systems” was launched in 1996, aiming at upgrading all its 402 large irrigation systems (each with irrigation area larger than 20 000 ha) by year 2015 through policy innovation, infrastructure rehabilitation, institutional reform and technical evolution. From 1996 to 2003, around US$4 billion have been invested jointly by the central government and local governments, 2 million ha of irrigation area have been rehabilitated or developed, and water services on 4.66 million ha of irrigation areas have been improved. Total grain production capability in these areas has increased by 11 million tonnes whereas total irrigation water consumption has reduced by 12 billion m³.

Malaysia emphasizes the agriculture sector as “the third engine of growth” in the national development agenda and has designated eight large irrigated rice production areas as the “rice granary areas” of Malaysia. In support of these, government investment for irrigation and drainage has boomed during the last decade and an irrigation modernization programme aiming at further improving irrigation infrastructures as well as addressing managerial, institutional and technological development has been formally established under the Seventh Malaysia Plan and is under implementation.

Table 4. Development expenditure for drainage and irrigation in Malaysia

Source: Country paper from Malaysia (in this publication).

In Lao PDR, Myanmar and Viet Nam, significant irrigation development is taking place with a variety of models: large-scale and small-scale, gravity and pumping, surface water and groundwater. The major focus has been on hardware construction. Software systems are also gaining attention, but somehow are gearing in at a slower pace. From 1997 to 2004, the irrigation area in Lao PDR increased by 140 percent through diversified water resources development, including weirs, reservoirs, pumping systems, diversion gates, traditional weirs and gabion dams. Some large reservoir systems are still at the design stage. In Myanmar, the most significant growth in irrigation, ever, happened during 1995 to 2005, with a net increase of about 700 000 ha, mostly in the form of reservoir systems. In Viet Nam, modern irrigation development stagnated until the reunification of the country in 1975. Early post-1975 growth was in small and medium irrigation schemes, whereas in the period 1985 to 1990, growth was concentrated on large irrigation and multipurpose schemes. The total irrigated area expanded at a rate of 2.9 percent/year in the period 1980 to 1987, whereas between 1988 and 1994 it was 4.58 percent/year. That trend is still continuing. Pumping irrigation plays an important role in Viet Nam and accounts for 26 percent of its total irrigation area.

Current status

From the early 1960s until the end of the twentieth century, the irrigated area in Southeast Asia doubled. The total irrigation area in this subregion reached about 18 million ha. Rice is the most important irrigated crop, accounting for 80 percent of all irrigated cropping areas. Taking southeast China into consideration, the percentage is about 75 percent. In Cambodia, rice represents 98 percent of irrigated crops. In Indonesia, Lao PDR, Malaysia, and the Philippines the percentages are higher than 80 percent.

In Malaysia, the Philippines and Thailand 102 large rice irrigation systems (larger than 10 000 ha) have been constructed, with about 3 million ha of total irrigation area, accounting for 44.4 percent of the total rice irrigation area in these three countries. In Indonesia, 3.2 million ha of rice field are irrigated by large rice irrigation systems (larger than 3 000 ha), accounting for 42 percent of the total rice irrigation area. China has 402 large-scale irrigation systems (larger than 20 000 ha), the total rice irrigation area accounts for 30 percent of the total national irrigation area (56 million ha); 5 300 medium-scale irrigation systems (between 667 ha to 20 000 ha) service an irrigation area representing 15.5 percent of the total national irrigation area. About 50 percent of these systems are rice-based. These large irrigation systems are the most important systems supporting food security and rural development. China concluded that its capacity to feed 22 percent of the world population with 9 percent of the world arable land was because it has 20 percent of world irrigation area, and the large-scale irrigation schemes are the bases of basic food production. Malaysia’s eight granary rice irrigation areas cover only 36 percent of the total paddy fields, but produce 72 percent of the total national rice production.

Table 5. Irrigation development in Lao PDR 1997–2004

Source: Mr Ouanpasith Chounlamany, Ministry of Agriculture and Forestry, Lao PDR, 2005, personal communication.

Table 6. Rice irrigation in Southeast Asia (and China) in 2000

Source: FAO AQUASTAT Database (2004).

Source: Country paper from Myanamar (in this publication).
Figure 1. Annual progress of reservoir water storage and irrigable area in Myanmar

Table 7. Large rice irrigation systems in some Southeast Asia countries (and China)

Note: Large irrigation system in Malaysia, Philippines and Thailand means larger than 10 000 ha.
         Large irrigation system in China means larger than 20 000 ha.
         Large irrigation system in Indonesia means larger than 3 000 ha.
Source: Country papers from Malaysia, Philippines and Thailand (in this publication).

Most of the large rice irrigation systems in Southeast Asia have been designed under a supply-driven mode. Despite their huge contribution to agriculture production and local social and economic development, there was a general consensus that these large rice irrigation systems had not lived up to expectations because of a legacy of poor institutional arrangement and system design, degraded infrastructure, poor management and stagnation in the face of rapid transformations of agriculture and pressure on their water supply. From 2000 to 2005, the FAO Regional Office for Asia and the Pacific conducted appraisals of a number of large rice irrigation systems in Indonesia, Malaysia, Philippines, Thailand and Viet Nam, and has identified some major constraints (Facon, 2005).

Institutional and administrative systems are generally top down. System planners, designers, managers and beneficiaries are not integrated in a systematic participatory approach. Design standards have not been changed for 20 to 30 years. Canal structure settings do not allow easy water control, measurement and system operation. Although most of these large irrigation systems were designed as multifunctional systems (municipalities, industrial customers), none of them has specific environmental targets. Recirculation of drainage is practised in a large number of schemes, but none is equipped with buffer or regulating reservoirs. Near-farm, and on-farm infrastructure is underdeveloped. System operation follows rigid seasonal schedules that lack flexibility. Field-level operators are poorly paid and their operation practices often differ from official rules and policies. Water supply from the main canals to the secondary canals and command areas is generally unreliable and inequitable. Effective monitoring and evaluation systems are not established or directly linked to operation. Managers do not have a proper estimation of system water balance and efficiency.

2. Future trends and requirements

2.1 Growing population calls for further development of irrigation

Southeast Asia is one of the most populated subregions, with about 522 million in 2002 and annual population growth of 2.1 percent. Nearly 33 percent of the total population lives in urban areas, and the urban population is further expected to increase by about 3 percent annually. Food security is still a major concern in this subregion. As indicated in Table 8, about 13 percent of the population in the developing countries of this subregion is undernourished. In Cambodia, Lao PDR, the Philippines, Thailand and Viet Nam, the proportions of the population undernourished are still above the global average. With the expected continuous population growth in the coming decades, food production in Southeast Asia needs to be increased. Irrigation is one of the major options to improve food security. Currently, irrigated agriculture is practised on 20 percent of all arable land and accounts for 40 percent of all crop production and almost 60 percent of cereal production in developing countries. FAO projects that from 2000 to 2030, to meet future food production demand, net global arable irrigated land needs to be expanded by 45 million ha. About 70 percent of the global expected increase in cereal production is attributed to irrigation. With abundant total water resources and lower development level, Southeast Asia is one of the subregions which have obvious irrigation development potential. For centuries, rice has been the most important source of food, employment and income for the rural people in Southeast Asia. Although rice consumption per capita is decreasing and is expected to decrease further, the special monsoon climate condition, the long heritage of social customs combined with the unique rice culture determine that rice is still the main staple food for the people of Southeast Asia and total demand will continue to increase. Rice irrigation in Southeast Asia is to be further expanded.

2.2 Increasing competition calls for more efficient water use

Southeast Asia has an average annual water resource of about 7 000 km³, about 15 percent of the world’s total. More than 90 percent of total freshwater withdrawals in the region go to agriculture, which is much higher than the global average of 70 percent. Currently, 18 percent of the harvested land in Southeast Asia is under irrigation. This rate is still low compared with the average level in Asia. Rice is a semi-aquatic plant, especially the lowland rice which accounts for about 90 percent of world rice production. The productivity of irrigated rice systems is threatened by competitive water use induced by rapid industrialization and urbanization. In China and Viet Nam, agriculture water withdrawal as a percentage of total national water withdrawal declined from 88.2 and 92.5, respectively, in 1990 to 67.7 and 68.1, respectively, in 2000. Overall, water competition from other sectors may not be a major concern in Southeast Asia, as actual consumptive use of water for irrigation is still expected to grow from 85.5 km³ in 1995 to 91.9 km³ in 2025, although more slowly than for other uses (Rosegrant et al., 2002). Nevertheless, in certain circumstances, increased competition for water from industry, domestic use and the environment will be a constraint for farmers, especially near urban areas and in drought years. This is dramatically changing the way we value and utilize water and the way we mobilize and manage water resources. It is necessary to produce more food and agricultural products with less water while maintaining the multiple roles of the systems. Figure 2 shows the example of Zhanghe Reservoir near Wuhan in Hubei province of China. Over a 30-years period, water allocated to agriculture from this reservoir has declined steadily from 80 percent to less than 20 percent. By implementing water saving practices at both system and farm level, there has been only a modest decline in agricultural production in the 100 000 ha Zhanghe Irrigation District.

Table 8. Undernourished population in Southeast Asia (and other world regions)

Source: FAO (2005).

2.3 Diversified agriculture calls for user-oriented water service

During the past 20 years, the importance of trade has increased rapidly in Southeast Asia. The international trade in agriculture (the average of agricultural imports and exports) increased from 47 percent of agricultural GDP in 1981–1983 to 89 percent in 2001–2003 (Dawe, 2005). Global cereal grain prices declined to 50 percent of their levels in the previous three decades and continued to a historical low in 2001, but there has been a comeback in Thailand in recent years. The agricultural sector in Southeast Asia has shifted from traditional rice cultivation to more diversified and market-oriented crop cultivation. In China, from 1990 to 2003, the rice crop area declined by 6.6 million ha (19.8 percent) and the vegetable crop area increased by 11.6 million ha (183.3 percent) and the orchard area increased by 4.3 million ha (82.2 percent). In Malaysia, the national policy is to decrease self-sufficiency in rice from 80 to 65 percent in 2010. Diversified agriculture demands increased flexibility and reliability in water services.

Table 9. Irrigation water withdrawal in Southeast Asian countries (and China) in 2000

Source: FAO AQUASTAT Database (2004).

Table 10. Urbanization rates in Southeast Asia: 1961, 1990, and 2004

Source of raw data: FAO (2005).

2.4 Demographic transition calls for new system innovations

Sustained economic growth and urbanization in Southeast Asia have been accompanied by significant demographic transition. Large numbers of farmers emigrated from rural areas to urban areas. Populations in some Southeast Asian countries, such as Indonesia, Malaysia and the Philippines have decreased in the past ten years. About 100 million farmers in China move to city areas seasonally for short term jobs. In addition to intersectoral mobility and growth in urban areas, rural household economies have become more composite. Farm households are diversifying their income sources outside of agriculture. In some areas, the household income from agriculture is even lower than that from non-agricultural occupations. In Central Luzon, Philippines, for example, the percentage of household income from agriculture declined from 64 percent in 1985 to 40 percent in 1997 (Hossain, 2000). As young males are migrating to urban areas, farm households are becoming older and more likely to be headed by females. In Thailand, agricultural census data show that the proportion of farm households headed by women increased from 12 percent in 1978 to 27 percent in 2003. The same source shows that the percentage of farm household heads aged 55 and above increased from 25 percent in 1978 to 34 percent in 2003. This situation makes it difficult to hire labour and causes the price of labour in rural areas to increase. For example, agricultural wages in the Philippines, adjusted for inflation, were 60 percent higher in 2002 than in 1981. This has significant impact on irrigation design, construction and management. New innovations based on labour savings, low cost, high productivity and operation and management ease need to be adopted.

2.5 Sustainable development calls for integrated water resources management

Competitive water use and segmented water resources management has resulted in rapid degradation of rural watersheds and the regional environment. In China, 70 percent of waterbodies of major rivers are polluted; in Indonesia, the number of degraded river catchments increased from 22 in 1984 to 59 in 1998; in Thailand and Viet Nam, mangroves and coastal crops have been destroyed as a result of contamination by shrimp farms. Rice irrigation is also responsible for some environmental issues such as non-point source pollution by nitrates and pesticides, overdraft of groundwater which has resulted in salt water intrusion into coastal aquifers and land subsidence and the gradual sinking of major cities such as Bangkok and Jakarta. Moreover, some 1.4 million ha of agriculture area are salinized by irrigation in Indonesia, the Philippines, Thailand and Viet Nam, and another 6.7 million in China. Other irrigation-related environmental issues include waterlogging, the release of acid (e.g., in the Mekong) and the spread of vector-borne diseases. It is therefore important to merge large rice irrigation management into integrated river basin or watershed water resources management.

Source: Barker and Molle (2004).
Figure 2. Annual water allocations, 1965–1999: Zhanghe Irrigation Reservoir, China

Source: Barker and Molle (2004).
Figure 3. Real world prices (1995 US$ per metric tonne) for rice, wheat, maize and urea

Table 11. Population change in Southeast Asian countries and China (per 1 000 inhabitants)

Source: FAO AQUASTAT Database (2004).

2.6 Ongoing democratization calls for participatory approaches

The planning and management of irrigation systems has been, and is being, shaped by the ongoing political processes of democratization, which constantly redefine the relationships between the state and the citizenry and have a bearing on the conditions of access to resources (Barker and Molle, 2004). The age of globalization thus brings with it a pressure to blend the traditional top-down decision-making by the state with the growing empowerment of civil society. This can be witnessed, for example, in the development of Asian non-governmental organizations (NGOs), some of which have successfully opposed some large-scale developments.

3. Measures undertaken in the past and present

3.1 Past route

Initial efforts at improving large irrigation system performance in Southeast Asia to a large extent have been concentrated on on-farm water management. The achievements have been very modest since they are normally partial, segmented, and have had little impact on the overall performance of the systems. Recently, governance strengthening and institutional reform have been high on government agendas. The achievements have been modest still mainly because farmers have limited capacity and few incentives to take over the operation and management responsibilities of irrigation systems that are in poor physical condition, and to pay for poor water services; also because the reforms have remained partial, with optimistic assumptions about the willingness or capacity of bureaucracies to carry out the necessary changes. More recently, technical concerns have made a comeback. Major system rehabilitation has been carried out for large irrigation schemes. But these works normally followed the original designs and concepts, focused on improving the performance of canal irrigation systems by lining canals, encouraged greater farmer participation, called for water pricing, cost recovery, and irrigation management transfer, neglected the changing trends and new requirements of the agriculture and water sectors, and neglected system design, operation and water services issues. At the 1996 FAO regional expert consultation meeting on modernization of irrigation systems, a new definition of modernization of irrigation systems was coined to guide future understanding and efforts, namely, “Irrigation modernization is a process of technical and managerial upgrading (as opposed to mere rehabilitation) of irrigation schemes combined with institutional reforms, if required, with the objective to improve resources utilization (labour, water, economic resources, environmental resources) and water delivery services to farmers.” Calls were made for systematic strategies to address institutional, physical and technical issues coherently through participatory approaches. Since then, numerous efforts have been taken and still are being taken by international communities and institutions and governments in Southeast Asia.

3.2 Recent responses and initiatives by FAO

FAO and the Land and Water Development Division have responded by developing, over recent years, in partnership with a number of international and national institutions, a series of technical and advocacy publications on the modernization of irrigation systems, tools for the appraisal and evaluation of the performance of irrigation systems, a suite of training materials and modules on the modernization of irrigation schemes, a regional training programme on irrigation modernization, a Website on the modernization of irrigation schemes, and has supported the efforts of national governments and agencies in the modernization of their irrigation sectors. The regional programme aims at disseminating modern concepts of service-oriented management of irrigation systems in member countries with a view to promoting the adoption of effective irrigation modernization strategies in support of agricultural modernization, improvement of water productivity, and integrated water resources management. The first training workshop under the programme was organized in Thailand in 2000, and since then India (Andhra Pradesh), Indonesia, Malaysia, Nepal, Pakistan, the Philippines, Thailand, Turkmenistan and Viet Nam, have benefited from the support of the regional training programme to organize national training workshops on irrigation modernization and benchmarking. More than 500 engineers and managers have been trained with support from the programme. The rapid appraisal process (RAP), introduced by FAO, has been adopted by the World Bank as one of the three elements of its holistic benchmarking methodology for irrigation systems.

3.3 Recent initiatives by governments and the international community

Although support from the international community in the form of investment in the irrigation sector has somewhat decreased in recent years, countries in Southeast Asia continue to have significant and ambitious programmes and objectives in the irrigation sector, particularly in terms of improving their large irrigation systems. For instance, the Royal Irrigation Department of Thailand has formulated a national strategy to improve irrigation efficiency and water management in existing systems while expanding new small-scale and medium-scale systems. Relevant activities on participatory management, conjunctive water use, water disaster mitigation and environmental protection have been initiated. A national training programme was also developed with support from FAO. The Department of Irrigation and Drainage (DID) of Malaysia is pursuing a national modernization strategy centred on the rice granary systems and has established a structured and elaborate programme to improve system performance and service quality. In Viet Nam, investment projects funded by the World Bank and the Asian Development Bank (ADB) include large irrigation modernization components based on similar concepts of service orientation. In China, the government has targeted that during the next 25 years, when national population grows from the current 1.3 billion to 1.6 billion, the agriculture sector should maintain national food security (95 percent self-sufficiency rate) with zero water consumption increase. Hence a nationwide water-saving programme is now under way based on legal, institutional, physical, technical and managerial options. The modernization of large rice irrigation systems is one of the core components and is now being implemented in more than 200 large schemes.

To summarize, all the efforts and initiatives that have been undertaken and that are being undertaken by government departments and the international community for the whole water sector have been focused on integrated water resources management and environmental issues through revising national water policies and strategies, establishing national water apex bodies and river basin organizations. For the irrigation subsector, the foci have been institutional reform through participatory irrigation management (PIM) and irrigation management transfer (IMT) to increase cost recovery and governance and adopting demand management to achieve improved performance through water pricing. Promotion of a private sector role and public-private partnerships in irrigation management and development have also been piloted. National policies have evolved from engineering expansion to food security, poverty alleviation, and related social objectives through multiple agriculture water use. An international programme on performance benchmarking in the irrigation and drainage sector is supported by FAO, the International Commission on Irrigation and Drainage (ICID), the International Programme for Technology and Research in Irrigation and Drainage (IPTRID), the International Water Management Institute (IWMI), and the World Bank.

3.4 Recent initiatives by rural communities and farmers

Farmers and system operators have adjusted to the challenges posed by the growing demand for water and new agricultural opportunities and constraints by exploiting groundwater, recycling water from drains and canals, changing cropping patterns, and adjusting the timing of water releases. These changes have taken advantage of new and cheap pumping technologies and government subsidies. Where system operations and management have not been proactively managing these changes, they have occurred nevertheless, with a growing dichotomy between official management and operating rules, and actual water management practices. Farmers still have little to say in general in the design and management of public irrigation schemes and in the definition of the service.

3.5 Outcomes and constraints

Some positive results have been achieved. In China, nationwide irrigation water use efficiency has been increased by around 10 percent during the past ten years. From 1980 to 2000, the total irrigation area in China increased by 6.7 million ha, the impoverished population in China reduced from 250 million to 29 million, while the total irrigation water use amount remained at about 350 billion m³ and the share of irrigation water use in total national water use declined from 85 percent to 63 percent. In Myanmar, irrigation development in lower Myanmar has enabled the rice cultivation area to increase from 4.78 million ha in 1988 to 6.54 million ha in 2003. Thus, rice exports increased to one million tons in 2004 (Naing, 2005). Despite these positive examples, overall progress on the 1996 modernization agenda has remained relatively modest. Constraints still remain. The concepts of irrigation modernization are not well understood and adopted. In some cases, it is just a resort to continue to obtain funding for rehabilitation, operation and maintenance, or further capital-intensive interventions. Policy changes have little impact since they are based on a poor understanding of basin and system efficiencies. Reformed institutions do not capture the complexity of the hydrological cycle and the multifunctionality of the irrigation systems and service relationships between different levels of management. In most countries, PIM/IMT has made very modest progress on improving system productivity and raising cost recovery rates. The appraisals of large irrigation systems in Southeast Asia confirm that, in general, the performance of these systems continues to be low in terms of control, water productivity, yields, and quality of service delivery to farmers. The differences between stated policies and actual practices are generally large. Significant underinvestment in operation and maintenance and poor management continue to be the norm rather than the exception. Very frequently the actual system performance, particularly service delivery, is overestimated and therefore they lack the capacity to support and enable the proposed reforms. On the other hand, the actual performance of the systems in terms of overall water use efficiency may have been considerably underestimated and therefore, the potential gains in water savings may have been considerably exaggerated (Facon, 2005).

4. Issues and options

4.1 Issues to be addressed urgently

Typological classification Southeast Asia is a diverse subregion including least-developed countries, developing countries and almost-developed countries. Different social and economic development levels pose different requirements for irrigation systems, hence the need to identify different models and approaches. No one model fits all. Whereas some countries are advocating multiple uses of rice irrigation systems, the least developed countries may still be focusing on food security. To address food security issues, rice irrigation needs to be expanded. Principally, expansion should be in those areas where land, water and labour resources provide comparative advantages for rice cultivation. But there is still a series of questions that need to be answered: what kind of approach should be adopted? Should it be existing system modernization or new system expansion or some combination of the two? For new system expansion, what kind of model should be adopted — large systems or small systems, surface water or groundwater or conjunctive use, gravity or pumping? These issues shall be addressed through typological classification.

Water productivity This is a prerequisite to justifying continuous water allocation to rice irrigation in an increasingly competitive water environment. Contrary to popular misconceptions, when grown under flooded conditions, rice has similar transpiration efficiency to other cereals. Lowland rice fields lose large amounts of water by seepage, percolation, run-off and evaporation from the water surface, and therefore require up to two to three times more water than other cereals. Much of these outflows is captured and reused downstream, and is not a true loss from rice-based systems. However, as widely discovered, wastage and misuse of water resources do exist in many rice irrigation systems and could be avoided. So this is involves improving water management and usage while renewing the notions and evaluation methods of water use efficiency and productivity.

Irrigation service Agriculture diversification and demographic transition call for more flexible, reliable and equitable irrigation services and low-cost, labour-saving operations. Large rice irrigation systems are designed and constructed for rice irrigation, and operated following rigid schedules. They are not compatible with diversified and quick shifting cropping patterns. Because of large-scale and complicated engineering settings, the requirements of maintenance work are high. This will need joint efforts to modify system design, reform institutions, rehabilitate infrastructure, manage innovation and build capacity.

Environmental protection The hard part is how to incorporate irrigation water management into integrated river basin water resources management to develop integrated strategies, systematic approaches and practical technologies to minimize their negative externalities while maintaining and further developing their positive externalities. Also, there may be a need to balance the needs of food security and multiple functions in the least developed countries.

Participatory management Large irrigation systems are normally designed for multiple uses with complicated facilities, and they cover large areas. They require higher technical and management qualifications for management staff and need broad coordination in system operation and management. Some systems cover several counties, provinces and even areas larger than a small country. For these systems, participatory management should have different modalities from those adopted in small community systems that are mostly advocated by the international community. Since farmers have limited capacity to manage and operate the main systems, and technical agencies and government may continue to play important roles in system management. Participatory approaches for democratic decision-making are needed, but suitable options need to be carefully identified.

4.2 Options proposed by experts and institutions

Policy innovation

• Align and harmonize water and irrigation policies with agricultural and environmental policies and integrate them into overall socio-economic development policies; better understand and recognize the multifunctionalities of rice irrigation systems.
• Incorporate concepts and principles of irrigation modernization into upgrading of existing irrigation systems and development of new irrigation systems.
• Target systems goals at the total productivity of rice-based livelihood systems while continuously highlighting poverty alleviation and food security.
• Develop strategic planning and management approaches with a service orientation to support irrigation modernization strategies.
• Adopt participatory planning and design processes to better incorporate farmers’ needs into system management goals.
• Further decentralize authorities and responsibilities from irrigation bureaucracies to system managers and farmers’ organizations to enhance the clarity of operation policies, ease the involvement of farmers or WUAs in decision-making, and facilitate the establishment of incentive structures for farmers and employees.
• Enforce appropriate policy and make institutional arrangements for conjunctive management of ground and surface water to enable rational resource planning.

Institutional reform

• Develop new frameworks that can manage the complexity of the hydrological cycle, the multiple roles of irrigation systems and deliver irrigation and drainage service to farmers in a responsive, accountable and efficient manner.
• Adopt business approaches to tailor the institution to deliver specific performance goals in addition to governance and representation goals, and to improve service orientation and accountability, to move towards decentralized management, and reflect the diversity of stakeholders and water uses.
• Change models of farmers’ organizations to professionalized institutions that can provide new ranges of delivery and other services and reduce transaction costs for farmers, as labour costs are increasing and labour and management shortages are to be further expected.
• Overhaul public management institutions by means of financial autonomy, incorporation, professionalization, public-private partnerships, privatization and transfer to farmers’ organizations.
• Create irrigation authorities’ accountability and water users’ responsibilities through the establishment of water rights for individuals or WUAs, or through suitable amendment of existing irrigation acts; develop service attitude of government agencies.
• Establish enabling legislation and enforcement capabilities, which would allow WUAs to operate in a businesses model.

Management improvement

• Develop systematic approaches and methods to incorporate irrigation system management into integrated river basin water resources management.
• Formulate explicit water management objectives for different levels (farm, irrigation system, river basin and national level).
• Develop a systematic approach, procedures and methods to improve system water control, water services equity, reliability and flexibility.

Financial support

• Mobilize public funds to support the transformation of systems and institutions towards more agile and better performing systems.
• Shift investment policy from being rice-centric to being rice-aware.
• Establish new financial instruments to cover not only O&M but also upgrading of management and infrastructure assets at all levels of agricultural water management.
• Provide investment to drainage, tools and data for decision-making at both national and system level.

Technical evolution

• Improve system design by adopting conjunctive use of surface and groundwater, recirculation of drainage, buffer reservoirs at appropriate levels in the systems, improved design of control structures, operation and ordering procedures, piping of near-farm delivery, proper drainage systems and intensification of irrigation system management.
• Revise irrigation design standards, conduct irrigation modernization capacity building.
• Conduct research on water-saving irrigation technologies, their real water savings, long-term sustainability and environmental impacts.
• Introduce information and control technology and software for remote monitoring of spills, drains, and flow rates at major offtakes as a basis for the establishment of feedback mechanisms, as well as for a better understanding of the water balance of the systems.
• Disseminate technologies and study new options for flexible canal lining.
• Develop methods and tools for volumetric measurement and pricing.
• Develop methods and tools for evaluation and pricing of multifunctionalities of rice irrigation systems.

International cooperation

• Disseminate technology and share information among line organizations, country institutions and governments, especially on interactions between design standards, operation strategies, service level and water pricing.
• Seek international assistance from funding agencies for technical evolution, institutional innovation and engineering construction and upgrading.

5. Conclusions

Most of the large rice irrigation systems in Southeast Asia have been designed for rice irrigation under a supply-driven mode. Despite their huge contribution to agriculture production and local social and economic development, there is a general consensus that these large rice irrigation systems have not lived up to expectations because of a legacy of poor institutional arrangements and system design, degraded infrastructure, poor management and stagnation in the face of rapid transformations of agriculture and pressures on their water supply.

Population growth, economic development, urbanization and globalization have been and will continue to be the main drivers of changes in the agriculture and water sectors. Southeast Asia is expected to be the subregion where the greatest urban growth will occur over the next thirty years. Agriculture water use is facing more competition from other sectors. In the meantime, diversification, commercialization and modernization of agriculture calls for more flexible, high-quality, user-oriented water services. Large rice irrigation systems are facing multiple challenges. The preferred option is irrigation modernization through participatory approaches.

Responses have been made and initiatives have been taken by national governments, the international community and farmers on various aspects of field water-management improvement, institutional reform and governance strengthening, irrigation rehabilitation and system modernization. Modest outcomes have been achieved, but some major constraints remain: the concepts of irrigation modernization are not fully understood and properly adopted; policy changes have little impact; PIM/IMT has made very modest progress; the performance of these systems continues to be low in terms of control, water productivity, yields, and quality of service delivery to farmers; the differences between stated policies and actual practices are generally large; significant underinvestment in operation and maintenance and poor management continue to be the norm rather than the exception.

To respond better to the previous shortcomings and to meet the new challenges, multiple options and systematic approaches are needed in terms of strategy, institutions, financing, technology and international cooperation to transform large rice irrigation systems in Southeast Asia from supply-driven to demand-driven responsive systems, to improve their water service in terms of reliability, equity and flexibility and multiple uses, to enable farmers to boost agricultural and water productivity, to be more responsive to market opportunities and to contribute to environmental sustainability.

References

Anon. 2002. Water for food and rural development. Proceedings of pre-symposium for the third world water forum (WWF3), 2002, Shiga, Japan.

Bakker, M., Barker, R., Meinzen-Dick, R. & Konradsen, F. 1999. Multiple uses of water in irrigated areas: A case study from Sri Lanka. SWIM paper 8, IWMI, Colombo, Sri Lanka.

Barker, R., Li, Y.H. & Tuong, T.P. 2001. Proceedings of an international workshop held in Wuhan, China 23–25 March 2001. IWMI, Colombo, Sri Lanka.

Barker, R. & Molle, F. 2004. Evolution of irrigation in South and Southeast Asia. Research Report 5, Colombo, Sri Lanka.

Chen, Z. & Facon, T. 2005. Multiple roles of agriculture water management system. Bangkok, FAO.

Dawe, D. 2005. Key trends affecting agricultural water resources management in Southeast Asia. Paper presented at the regional workshop on the future of large rice-based irrigation systems in Southeast Asia, Ho Chi Minh City, Viet Nam, 26–28 October 2005.

Domingo, P. T. 2005. Development and management of large rice-based irrigation systems: Philippine scenario. Country paper for regional workshop on the future of large rice-based irrigation systems in Southeast Asia, October 2005, Ho Chi Minh City, Viet Nam.

FAO. 2005. The state of food insecurity in the world. Rome.

Facon, T. 2005. A rapid appraisal procedure to assess the performance of irrigation systems: lessons from an FAO regional irrigation modernization and management training programme in Asia. Bangkok, FAO.

Groenfeldt, D. 2005. Multiple roles of irrigated agriculture and investment implications. Background paper for the Comprehensive assessment of water management in agriculture, Santa Fe, New Mexico.

Guerra, L.C., Bhuiyan, S.I., Tuon, T.P. & Barker, R. 1998. Producing more rice with less water from irrigated systems. SWIM paper 5, IWMI, Colombo, Sri Lanka.

Huang, Y., Wang, H., Huang, H., Feng, Z.W., Yang, Z.H. & Luo, Y.C. 2005. Characteristics of methane emission from wetland rice-duck complex ecosystems. Agriculture Ecosystems and Environment 105 (2005) 181–193.

Kristianto, D. 2005. Present performance on large rice-based irrigation systems in Indonesia. Country paper for regional workshop on the future of large rice-based irrigation systems in Southeast Asia, October 2005, Ho Chi Minh City, Viet Nam.

Maeda, T. 2004. Bird use of rice field strips of varying width in the Kanto plain of central Japan. Agriculture Ecosystems and Environment, 105: 347–351, Morika, Japan.

Matsuno, Y., Ko, H.S., Tan, C.H., Barker, R. & Levine, G. 2002. Accounting of agriculture and non-agricultural impacts of irrigation and drainage systems. Working paper 43, IWMI, Colombo, Sri Lanka.

Meusch, E., Yhoung-Aree, J., Friend, R. & Funge-Smith, S. 2003. The role and nutritional value of aquatic resources in the livelihoods of rural people. Bangkok, FAO.

Naing, M. 2005. Paddy field irrigation systems in Myanmar. Paper presented at the regional workshop on the future of large rice-based irrigation systems in Southeast Asia, Ho Chi Minh City, Viet Nam, 26–28 October 2005.

Renault, D. & Facon, T. 2004. Beyond drops for crops — the system approach for water value assessment in rice-based production systems. Rome, FAO.

Renault, D. & Montginoul, M. 2002. Positive externalities and water service management in rice-based irrigation systems of the humid tropics. Agriculture water management, 59: 171–189, 2003.

Rosegrant, M.W., Cai, X. & Cline, S.A. 2002. World Water and Food to 2025. International Food Policy Research Institute, Washington, DC.

Taniyama, S. 2002. Water resources and rice paddy cultivation in the Asia monsoon region. Tokyo, Japanese National Committee of ICID.

Ti, L.H. & Facon, T. 2004. From vision to action — a synthesis of experiences in least-developed countries in southeast Asia. Bangkok, FAO.

Tomith, T.P., Thomas, D.E. & van Noordwijk, M. 2004. Environmental services and land use change in Southeast Asia: from recognition to regulation or reward? Agriculture ecosystems and environment, 104: 229–244, Morika, Japan.

Turner, K., Georgious, S., Clark, R., Brouwer, R. & Burke, J. 2004. Economic valuation of water resources in agriculture, FAO water reports 27, Rome.

UNESCO. 2003. World water development report, 2003. Water for people, water for life. UNESCO-WWAP, New York and Oxford, Berghahn Books.

¹ Water Resources Development and Conservation Officer, FAO Regional Office for Asia and the Pacific, Bangkok, Thailand; [email protected]

Contributors²: Atlin G.N., Balasubramanian V., Barry G., Bennett J., Dawe D., Dittert K., Facon T., Fujimoto N., Gupta R.K., Haefele S.M., Heong K.L., Hosen Y., Ismail A.M., Johnson D., Johnson S., Khan S., Lin Shan, Masih I., Matsuno Y., Pandey S., Paris T.R., Peng S., Thiyagarajan T.M., Wassman R.

Executive summary and key messages

Overall message. To sustain the livelihoods of rice farmers and meet the food needs of growing populations in the face of increasing water scarcity and abiotic stresses, low cost technologies that increase production and water productivity at field and system scales, and that minimize negative environmental externalities and maintain ecosystem services, need to be developed and adopted.

Rice is the most important food and biggest agricultural water user in the world. Dramatic changes are occurring in the biophysical and socio-economic environment under which rice is produced. In the face of increasing resource constraints — land, labour, and water — new technologies and management practices must continue to be adopted to sustain the livelihood of rice farmers in irrigated, rainfed, and unfavourable ecosystems and to meet the food needs of a growing population. Particular attention must be given to practices that increase water productivity and protect the environment. We need a better understanding of the interactions between field, irrigation system, and landscape level and of the various ecosystem services of rice-based landscapes to meet both agricultural and environmental development objectives.

Submessage 1. Rice is the main staple food for three billion people, 90 percent of whom live in Asia.

For centuries, rice has been the most important source of food, employment and income for the rural people in Asia. The concept of a rice culture gives meaning to rice beyond its role as a food item. Increases in rice production over the last 50 years have played a major role in achieving food security and alleviating malnutrition and poverty, especially through lowering the rice price. However, the green revolution has largely bypassed the unfavourable drought-, flood-, and salinity-prone areas where nearly 700 million people depend on rice production. Sustained rapid economic growth and urbanization in Asia will continue to increase the feminization of agriculture and reduce the availability of labour for rice cultivation, driving the need for high-yielding production technologies with lower labor requirement.

Submessage 2. Rice grows in a wide range of environmental conditions and is productive in many situations where other crops would fail.

Rice is unique in its ability to grow and yield in a wide range of agro-ecological conditions, from flooded lowlands to drought prone uplands, and from humid tropical to cool temperate climates. Although rice is adapted to water logging, deep water decreases productivity and complete submergence can be lethal. Rice is a semi-aquatic plant and yield declines as the soil dries below saturation. Therefore, rice is grown under ponded water culture (lowland rice) where possible, accounting for about 90 percent of world rice production. Contrary to popular misconceptions, when grown under flooded conditions, rice has similar transpiration efficiency to other cereals. Nonetheless, lowland rice fields lose large amounts of water by seepage, percolation, runoff and evaporation from the water surface, and therefore require up to two to three times more water than other cereals. However, much of these outflows is captured and reused downstream, and is not a true loss from rice-based systems.

Submessage 3. The productivity of irrigated rice systems is threatened by increasing water scarcity, whereas that of rainfed and drought- and flood-prone environments is affected by a complex of abiotic stresses.

The highly-productive irrigated rice ecosystem receives 17 to 28 percent of the world’s developed fresh water resources and provides cheap food for urban and rural landless rice consumers. Many large-scale irrigation systems are in poor condition and poorly managed. Groundwater pumping has increased tremendously relative to surface irrigation and overdraft has become a serious problem in the semi-arid regions. Water scarcity is increasing in some regions because of competition with urban and industrial users, and because of climate change. There is no systematic inventory or quantification of the nature, extent or severity of water scarcity in rice-growing areas, and its likely impact on productivity. The rainfed and unfavourable rice ecosystems experience multiple abiotic stresses such as drought, salinity/sodicity and uncontrolled flooding.

Submessage 4. Lowland rice ecosystems have both beneficial and negative environmental externalities and unique ecosystem services (multifunctionality) which will be affected by increasing water scarcity.

Compared with other cereals, lowland rice is a heavy emitter of methane and ammonia and a low emitter of nitrous oxide. However, lowland rice fields are responsible for less than 10 percent of total global methane emissions. Lowland rice fields behave as artificial wetlands in their capacity to remove nitrogen and phosphorus from contaminated surface waters. Nitrate leaching from flooded rice fields is usually negligible. Compared with temperate non-flooded crops, biocide use in Asian rice systems is generally low and the biocides used degrade rapidly. However, the biocides are often extremely toxic and their negative human health impact is large. Flooded rice can increase the risk of salinization and water logging in poorly drained areas by raising groundwater levels. Conversely, it can be used to reclaim saline and saline/sodic soils by leaching salts from the rootzone. Often overlooked are non-food ecosystem services provided by the lowland rice landscape, such as cultural aspects, groundwater recharge, control of soil erosion, flood mitigation and sustenance of a rich biodiversity, including unique and endangered species. The multiple uses and functions of rice landscapes need to be valued.

Submessage 5. More rice needs to be produced, at low cost, in the face of increasing water scarcity and a complex of abiotic stresses, while protecting the environment and safeguarding the ecosystem services (multifunctions) of rice-based ecosystems.

Meeting future rice demand and reducing poverty will require: increasing resource productivity in irrigated, rainfed and unfavourable ecosystems; reducing production costs and labour requirements; improving the management of water resources in the face of declining availability; and development of technologies and strategies to cope with the likely impacts of climate change, including increased incidence of extreme events. Increasing water scarcity in many areas will change the dominantly flooded systems to more aerobic systems, bringing new challenges for increasing productivity while minimizing associated negative externalities and maintaining beneficial ecosystem services (multifunctionality) of rice-based systems.

Submessage 6. There is scope to greatly increase rice productivity in unfavourable regions by developing rice varieties that are drought-, salt- or submergence-resistant, but there is little scope to further increase yield potential of current high-yielding (inbred or hybrid) varieties grown under non-stressed conditions.

Yield potential of modern high-yielding varieties has stagnated during the last two decades, with the exception of the development of hybrid rice. Transforming rice from C3 into C4 may increase the yield potential, but there is no consensus about its feasibility. Theoretically, there are a range of options to increase the water productivity of rice with respect to evapotranspiration by manipulating early crop vigour, leaf waxiness and transpiration efficiency, but these have not yet been explored. Traits that improve nutritious quality need to be combined with other successful traits, including those that impart drought tolerance. With proper investments, varieties can be developed with 50 to 100 percent increase in yield potential in rainfed lowlands and drought- and flood-prone areas within ten years. Genetic engineering tools can accelerate the development of these new varieties. Management technologies will also need to be developed and adopted to achieve the potential gains offered by the new varieties.

Submessage 7. There is great scope to develop and deploy integrated technologies to increase rice productivity and lower production costs in the face of water scarcity, but more research is needed on real water savings, long-term sustainability and environmental impacts.

Technologies that integrate components of management with varietal improvement can bridge the yield gap in well-defined target environments. The water balance of rice fields under such technologies should be quantified to identify water savings at field and system scales. Several water saving technologies are being developed for water-short irrigated environments which need to be further explored in participatory research. Integrated development of varieties and management practices is required to optimize yields in these environments. The sustainability and environmental impacts of many newly developed technologies are not well understood. In rainfed and drought- and flood-prone environments, technologies should aim at reducing abiotic stress intensities, enhancing survival and robustness of the crop to withstand stress, and stabilizing yields.

Submessage 8. More understanding is needed of scale effects and rice ecosystem services (multifunctionality) to develop specific improvement options, especially for irrigated ecosystems.

Many generic options to improve irrigation systems or the productivity of rainfed and unfavourable agro-ecosystems apply to rice-based systems as well. However, irrigated rice systems are characterized by large volumes of water flows, and more insight is needed on current water reuse and on system-level impacts of adoption of field-level water-saving technologies. More systematic evaluations are needed to understand the water balance dynamics and productivity of water at various scales of the irrigation system. More case studies are needed to identify local or region-specific characteristics of ecosystem services. Management practices need to be developed to sustain and enhance the ecosystem services of rice landscapes.

Submessage 9. Appropriate incentives and policies are needed to steer the improvement of water use in rice-based ecosystems.

Proposed policies, institutions and incentives to increase the efficiency and equity of water use in agriculture in general apply to rice-based systems as well. These include water pricing, tradable water rights, water markets, virtual trade, increasing the role of the private sector in operation and management, and general rural development strategies. A particular characteristic of rice-based cropping systems is that holdings are small and that some policies, such as volumetric pricing, should better be targeted at groups of farmers rather than at individual farmers. With the growing scarcity of labour and water, the development and dissemination of labour- and water-saving practices and drought-tolerant rice varieties are priority areas for research.

Trends and conditions

1. The role of rice in food security and poverty alleviation

Rice production

Rice is eaten by about three billion people and is the staple food of the largest number of people on earth. Ninety percent of the world’s rice is produced and consumed in Asia, where it accounts for more than half of the agricultural harvested area in many countries and for 20 to 40 percent of total caloric intake. An important political objective in most rice-growing countries is to achieve self-sufficiency in rice production and maintain price stability through domestic procurement and adjustment of stocks. The international trade in rice is only some 5 percent compared with 18 percent for wheat and 11 percent for maize. The growth in rice production over the last 40 years has kept pace with the tremendous growth in population in Asia and now stands at about 550 to 600 million tonnes annually. The average per capita rice production is about 18 percent higher than it was in the mid-1960s. This increase in rice production was largely a result of expansion of the irrigated area and increase in yield (t ha^-1) of irrigated rice.

Rice production and poverty alleviation

The surplus rice produced in the irrigated systems contributes to food security and helps to alleviate the poverty of an increasing urban population. The increased production from irrigated rice (the green revolution) was the result of a combination of the development of high-yielding varieties, expansion of the irrigated area and increased inputs such as water, fertilizer, and pesticides. High yield potential was achieved through increased harvest index, dwarfing and lodging resistance. Over the years, short growth duration (allowing multiple-cropping) and resistance to major pests and diseases were incorporated into the high-yielding varieties. Access to irrigation substantially reduced the probability of living in poverty. However, the main driving force in poverty reduction was the decline in the rice price that resulted from the increase in rice production. The world rice price now is just 25 percent of its level in the early 1980s and national prices show a similar trend. Rice farmers have borne the brunt of lower prices whereas rice consumers, particularly low income consumers, have been the major beneficiaries. In particular, the urban labouring class, the rural landless, and the small and marginal producers (being net consumers) benefit from the low rice price.

Though the green revolution was successful in dramatically increasing yields in favourable, irrigated ecosystems, it largely bypassed the unfavourable drought-, flood-, and salinity-prone areas. Nearly 700 million people are dependent on rice production in these environments, mainly in South and Southeast Asia. Countries and regions with a large proportion of rainfed areas have a relatively high incidence of poverty. These rainfed areas now constitute the major hunger and poverty hotspots of the world. Raising the productivity of major staples such as rice is critical to addressing hunger and poverty in these areas.

The Asian rice culture and economy

For centuries, rice has been the main staple food in most parts of Asia and has been the most important source of employment and income for rural people. It is currently estimated that rice is grown on some 250 million farms in Asia, most of them family-owned with average size varying from less than 0.5 to 4 ha. Much rice is used for home consumption and the marketed ratio for rice is small. In Asian countries, the rice production system is also highly valued because of its strong links to culture, rural communities, natural resources and the environment. Rice affects daily life in many ways and the concept of a rice culture gives meaning to rice beyond its role as an item of production and consumption. There are many traditional festivals and religious practices associated with rice cultivation. Many old kingdoms as well as small communities have been founded on the construction of irrigation facilities to stabilize rice production. The decline in these hydraulic societies is often associated with the inability to manage water resources.

As of the last century, Asia is becoming more urbanized. At the same time, the distinction between rural and urban is becoming blurred and households increasingly have representation in both worlds. In 1950, 17 percent of the Asian population was urban. This share is now 40 percent and will exceed 50 percent by 2025. With the decline in the rice price, farmers who once depended largely on their income from rice now earn half or more of their income from non-rice sources, including non-farm sources. For example, between 1985 and 1997, among a sample of Central Luzon rice farmers, income from rice declined from 48 to 35 percent. A similar decline occurred in Thailand.

As economies develop, the rural sector undergoes major changes. The younger members, particularly the males, leave rural areas in search of jobs in urban areas or overseas, sending remittances to their rural homes. In many instances the rural economy is becoming more feminized and aging. The feminization of agriculture is likely to continue to increase.

Labor demand from the expanded industrial and service sectors has increased wages. Farm employment is less desirable and it is harder to find labour at peak periods for key operations such as transplanting, weeding, and harvesting. With rising wages and labour shortages, mechanization is becoming more prevalent for both land preparation and harvesting, especially in the irrigated areas. Farmers are shifting from hand weeding to the use of herbicides, and from transplanting to direct seeding. As of the late 1990s it is estimated that one-fifth of the area in Asia is direct seeded, and it is expected that this proportion will increase in the future.

2. Rice ecology and hydrology

Rice ecology

Rice is unique among the major food crops in its ability to grow in a wide range of hydrological situations, soil types and climates. Depending on the hydrology of where rice is grown, the rice environment can be classified broadly into irrigated rice, rainfed or unfavourable ecosystems. Irrigated lowland rice is grown in bunded fields with assured irrigation for one or more crops per year. Usually, farmers try to maintain 5 to 10 cm of water (floodwater) on the field. Worldwide, there are about 79 million ha of irrigated lowland rice. Rainfed lowland rice is grown in bunded fields that are flooded with rainwater for at least part of the cropping season to water depths that exceed 100 cm for no more than ten days. Worldwide, there are about 54 million ha of rainfed lowland rice. In both irrigated and rainfed lowlands, fields are predominantly puddled with transplanting as the conventional method of crop establishment. In flood-prone ecosystems, the fields suffer periodically from excess water and uncontrolled, deep flooding. About 11 to 14 million ha worldwide are flood-prone lowlands. Upland rice is grown under dryland conditions without irrigation and without puddling, and is grown on some 14 million ha.

In many rice production areas, rice is grown as a monoculture with two crops per year. However, significant areas of rice are also grown in rotation with a range of non-rice crops, including about 15 to 20 million ha of rice-wheat systems, much of which is irrigated. In such systems, the preferred cultural practices for rice are often antagonistic to the preferred practices for the non-rice crops, and vice versa.

The rice plant

Cultivated rice evolved from a semi-aquatic, perennial ancestor, and rice evolved separately from other Graminae before grasses moved from the forest floor to more open habitats. The wetland ancestry of rice is reflected in a number of morphological and physiological characteristics that are unique among crop species. Key differences between rice and other cereals include shoot and root anatomy, water loss patterns, and growth responses to soil water status drier than saturation.

Rice is extremely sensitive to water shortage. When soil water content drops below saturation, growth and yield formation are affected, mainly through reduced leaf surface area, photosynthesis rate and sink size. In rainfed systems, drought stress can occur at any time during the growing season, and is especially damaging immediately before and during flowering. There is little relationship between tolerance to stress at the vegetative and reproductive stages. Genetic analysis of traits related to drought tolerance in rice indicate that grain yield under drought stress in rice is a complex trait affected by many genes. No genes for tolerance to either reproductive- or vegetative-stage stress with effects large enough to be useful in breeding have yet been identified. Screening for drought tolerance is complicated by the intermittent occurrence of natural stress, and by the strong relationship between plant phenology and sensitivity to stress.

Although rice is adapted to water logging, complete submergence can be lethal. The effect of flash-flooding varies depending on growth stage. Germination is highly sensitive to flooding. Short-term flooding hastens the plant’s energy depletion and increases mortality. Tall plants tend to lodge when the water level recedes resulting in additional yield losses and poor grain quality. Under long-term flooding, genotypes need to be able to elongate fast to keep their leaves exposed above water surface. Productivity of this system is low because of the high energy required for shoot elongation. Most existing rice cultivars are seriously damaged if they are completely submerged for more than three days. Progress through breeding has been slow because of the complexity of tolerance where many traits are involved, and the complexity of the environment where more than one stress is prevailing.

Rice is a salt-sensitive crop with a threshold of about 2 to 3 dS m^-1. Salt stress affects rice through osmotic stress, salt toxicity and nutrient imbalances. Tolerance to salt stress in rice is complex and varies with the stage of development. Rice is most sensitive to salinity during the reproductive and early seedling stages. Tolerance at one stage does not correlate with tolerance at another stage, and traits essential for tolerance at one stage are different from those at another stage. These suggest the need for incorporating traits/genes associated with tolerance at different stages. Because of this complexity and the many traits involved, progress in breeding for salt tolerance has been relatively slow. Also, salinity typically occurs in association with other soil-related stresses or temporary tidal submergence.

Water use and water productivity

About 90 percent of the world’s rice production is harvested from so-called lowland (or paddy) rice fields. Traditionally, lowland rice is raised in a seedbed and then transplanted into a main field which is kept under continuous or intermittent ponded water conditions. Land preparation consists of soaking, ploughing and puddling (i.e., harrowing or rotivating under shallow submerged conditions). Puddling is done for weed control and to reduce soil permeability and percolation losses, and eases field leveling and transplanting.

Water for lowland rice is needed for land preparation and to match seepage, percolation and evapotranspiration outflows during crop growth. The soil is ponded for soaking prior to puddling and transplanting, and is usually kept ponded until shortly before harvest. Seepage is the lateral subsurface flow of water and percolation is the down flow of water below the rootzone. Typical combined values for seepage and percolation vary from 1 to 5 mm d^-1 in heavy clay soils to 25 to 30 mm d^-1 in sandy and sandy loam soils. Evaporation occurs from the ponded water layer or from the surface of the soil and transpiration is the process by which plants take up water from the soil and release it into the air as vapour. Typical combined evaporation and transpiration rates of rice fields in the tropics are 4 to 5 mm d^-1 in the wet season and 6 to 7 mm d^-1 in the dry season, but can be as high as 15 mm d^-1 in subtropical regions prior to the onset of the monsoon. Over-bund flow (or surface runoff) is the spillover when water depths rise above the paddy bunds.

Transpiration rates of lowland rice crops are the same order of magnitude as many other cereals, but the evaporation flows from soil or the ponded water layer are higher. The modern rice varieties, when grown under well-watered conditions, can have similar transpiration efficiency (water productivity with respect to transpiration) to other C3 cereals such as wheat, at about 2 kg grain m^-3 water transpired.

The few available data indicate that water productivity of rice with respect to evapotranspiration is similar to that of wheat, ranging from 0.6 to 1.6 kg grain m^-3 of evapotranspired water, mean 1.1 kg grain m^-3. For maize, being a C4 crop, the values tend to be higher, ranging from 1.1 to 2.7 kg grain m^-3 water, mean 1.8 kg grain m^-3. However, total water inputs to rice fields (rainfall plus irrigation) are up to two to three times more than for other cereals. Total seasonal water inputs to rice fields vary from as little as 400 to 600 mm in heavy clay soils with shallow groundwater tables, to more than 2 000 mm in coarse textured (sandy or loamy) soils with deep groundwater tables. Around 1 300 mm seems to be a typical average value for irrigated rice in Asia. Values of water productivity with respect to total water input range from 0.2 to 1.2 kg grain m^-3 water, with 0.4 as average value. Non-productive outflows of water by run-off, seepage and percolation are about 25 to 50 percent of all water input in heavy soils with shallow water tables, and 50 to 85 percent in coarse textured soils with deep water tables. Though seepage and percolation are losses at the field level, they are often captured and re-used downstream and do not necessarily lead to true water depletion at the irrigation area or basin scales. The proportion and magnitude of unrecoverable losses by seepage and percolation from rice fields is generally not known.

3. Rice ecosystems

The irrigated ecosystem

Irrigated rice provides 75 percent of the world’s rice production. The green revolution led to a rapid increase in productivity especially in irrigated areas during the 1960–80s. Country-average irrigated rice yields in Asia now range from 3 to 9 t ha^-1, with an overall average of about 5 t ha^-1. Irrigated rice receives 17 to 28 percent of the world’s developed fresh water resource. Approximately 56 percent of the total world irrigated area is in Asia, where rice accounts for 40 to 46 percent of the net irrigated area. In Southeast Asia, rice occupies 64 to 83 percent of the irrigated area, in East Asia 46 to 52 percent, and in South Asia 30 to 35 percent. Irrigated rice is mostly grown with supplementary irrigation in the wet season, and is entirely reliant on irrigation in the dry season. Irrigated wet-season rice is grown predominantly in the subtropical regions of north and central China and the northwest Indo-Gangetic Plain of India and Pakistan, and is highly dependent on irrigation. Dry season irrigated rice is concentrated in south China, south and east India and Bangladesh. The proportion of the rice area that is irrigated (excluding China where essentially all rice is irrigated) increased substantially from the late l970s (35 percent) to the mid-1990s (44 percent). This occurred because of an increase in the irrigated area coupled with a large decline in upland and deep water rice cultivation. The growth in the total irrigated area in Asia has slowed in the past decade and is projected to increase at less than 1 percent per year between 1995 and 2020. Lower rice prices, poor performance and high construction costs of irrigation systems, and environmental protests over large dam construction are major factors behind this decline in growth rate. Large parts of the increase in irrigated area of Asia since 1980 are not being planted to rice. For example, 45 percent of the gross irrigated area in India was cropped to rice in 1960, but by 1992 this had fallen to 30 percent.

Irrigated rice production has only recently moved into non-traditional rice growing areas, on relatively permeable soils, in the northwest Indo-Gangetic Plain. Between the middle of 1960 and the late 1980s, the area, yield and total production of rice increased rapidly as a result of the introduction of high yielding varieties, increased use of chemical fertilizers, assured irrigation via public investment in irrigation systems and private investment in groundwater pumping, and favourable policies (subsidies for inputs and minimum price support schemes for grain). However, there are now grave concerns about the sustainability of irrigated rice production at current levels in this region because of rapidly falling groundwater levels and the need to reduce the large fiscal costs associated with government policies that promote rice production.

Shifting comparative advantage in rice production

The comparative advantage in rice production is shifting within Asia. Before World War II, the delta regions (Bangladesh, Cambodia, eastern India, Myanmar, Thailand, Viet Nam) held the comparative advantage in rice production and were the main sources of rice exports. However, the early beneficiaries of the green revolution technology were those areas where, with the construction of reservoir storage, it was possible to irrigate two crops of rice, and the northwest Indo-Gangetic Plain where private groundwater pumping expanded rapidly. For political reasons and/or because of the inability to manage floods, the deltas initially were unable to take advantage of the new rice technologies. However, over the past 15 to 20 years, the delta areas have regained the comparative advantage with the aid of low-cost pump technology. During this period, the delta regions have shown the most rapid growth in rice production and exports. With the improved water control provided by pumps, the delta regions have been able to shift out of deep water and floating rice by planting one crop before and one after the floods. In short, rice production is gaining in those regions with plentiful water supply and cheap labour relative to those areas of water scarcity.

Irrigation water supply infrastructure

The growing role of the private sector in managing ground and surface water irrigation is changing the incentives to manage water. Much of the existing surface irrigation infrastructure in Asia has been designed for supplementary irrigation of rice during the rainy season. Large systems are publicly managed in a supply-driven mode without accountability to users. They are generally in a poor condition because of insufficient maintenance, and provide poor control and service to farmers. Design standards and operation have not changed in many countries for 20 to 30 years: they are a cause of poor performance and have not responded to changes in farming systems and farmers’ service requirements, and their extension into the dry season. In gravity systems water is either priced well below operation and maintenance costs or not priced at all. The general response from the public sector has been their rehabilitation, transfer of the burden of maintenance costs to farmers, and substantial investments in rigid canal lining, with dubious results.

A very important trend is the tremendous increase in irrigation with groundwater pumping in the semi-arid regions of Asia since the 1960s (e.g. the northwest Indo-Gangetic Plain and the North Central Plain of China). More recently, there has been a rapid growth in the use of pumps in the monsoon areas. The use of pumps has facilitated diversification from rice to higher value crops. With privately owned systems the users pay for the cost of fuel and maintenance with the exception that electricity is subsidized in some areas (such as the Indo-Gangetic Plain). Users who pay the cost of pumping have an incentive to minimize costs by managing water carefully.

Water scarcity in the irrigated ecosystem

Though water scarcity is increasing worldwide, there is no systematic inventory, definition or quantification of water scarcity in rice growing areas. It has been estimated that by 2025, 15 to 20 million ha of irrigated rice will suffer some degree of water scarcity. Water scarcity is most pronounced in the semi arid regions of the Indian subcontinent where 15 000 million ha of rice are grown mostly in the wet season. In parts of the northwest Indo-Gangetic Plain and the Yellow River Plain, long surplus producers of cereals, heavy pumping of groundwater is rapidly lowering water tables. Even in many monsoon areas such as India’s Cauvary Delta and Thailand’s Chao Phraya Delta, there is heavy competition for water among states and sectors, particularly in the dry season. Individual case studies also suggest local hotspots of water scarcity, even in areas generally not considered water scarce.

Two typical cases of water scarcity can be defined. In some irrigated rice areas, there may be sufficient water for farmers to practise flooded rice culture. Tail end parts of irrigation systems or regions with falling groundwater tables are examples of areas where farmers may not have access to sufficient water to grow continuously flooded rice. The challenge in this case is to devise strategies to assist farmers to cope with water scarcity. In the second case, there may be adequate water where the farmer is (e.g. upstream part of an irrigation system), but inadequate water elsewhere in the system or the catchment, or there may be increasing demands from non-agricultural sectors of society (e.g. cities, industry) that result in partial or complete withdrawal of water resources from irrigation systems. In this case, the challenge is to have strategies for farmers to reduce water input, to “save” water to divert it to somewhere else or for use by other sectors.

Coping with water scarcity or reducing water inputs will require changing from flooded to more aerobic rice cultural systems and to irrigation systems that are more responsive to the needs of farmers. Changed water management will affect many aspects of crop management, productivity, and the environmental impacts of rice-based systems.

The rainfed and unfavourable ecosystems

The rainfed and unfavourable rice ecosystems (including upland, flood-prone and saline environments) account for about 45 percent of the total rice area and produce 25 percent of the world’s rice. The rainfed and unfavourable rice ecosystems experience multiple abiotic stresses. Approximately 25 million ha of rainfed rice are frequently affected by drought, the largest, most frequently and severely affected areas being eastern India (about 20 million ha) and northeastern Thailand and Lao PDR (7 million ha). About 15 million ha of rice land are frequently damaged by submergence in South and Southeast Asia and tropical Africa. Further constraints arise from the widespread incidence of problem soils. Salinity and/or sodicity are most widespread (>10 million ha) in coastal areas, where salinity is predominant, and in inland areas, where both salinity and alkalinity are major problems and often coexist.

The abiotic stresses of rice-based rainfed lowlands are characterized by high levels of uncertainty, particularly for timing, duration and intensity because of the unpredictability of rainfall and/or flooding. Most rainfed lowlands are characterized by small to medium topographic differences that have important consequences for water availability, soil fertility, flooding risk, and extent of soil-borne abiotic stresses. Most of the management options for rice were developed for irrigated ecosystems with a reliable supply of water. The unpredictability of rainfall in rainfed ecosystems often results in field conditions that are too dry or too wet. Besides imposing abiotic stresses on crop growth, these conditions also prevent timely and effective critical management operations such as land preparation, transplanting, weed control, and fertilizer application. If such operations are delayed or skipped, large yield losses often ensue, even though the plants have not suffered physiological water stress.

Because of diverse stresses, productivity growth in rainfed areas has been slow and average rice yield is currently some 2 t ha^-1. Research efforts to increase yields and yield stability in rainfed environments have been limited; specific research issues for rainfed rice have been extensively discussed only in the recent past. Thanks to these recent efforts, rice research for rainfed lowlands has been intensified and was more successful during the last decade. Together with socio-economic developments, this has contributed to considerable improvements in rainfed systems. Important changes include much-improved access to markets for inputs (inorganic fertilizer, agrochemicals), more opportunities for off-farm income, improved varieties, (partial) mechanization, and better access to information.

4. Environmental conditions

Air pollution

Rice stubble burning causes serious air pollution, affecting human health, in the mechanized rice-wheat systems of the northwest Indo-Gangetic Plain where the majority of the rice is harvested by combine harvester. Approximately 10 million tonnes of rice straw are burnt in the small state of Punjab, India, alone. Burning is promoted by the lack of economic uses for rice straw, the loss of wheat yield with delayed planting, and the lack of suitable machinery for direct drilling into rice stubbles. In contrast, burning of wheat straw is a major source of air pollution in rice-wheat systems in China where the turn around time between wheat harvest and rice planting is small. Policies banning stubble burning were recently introduced in Punjab, India and Sichuan, China, although the policies have not yet been legislated in India, and implementation will be difficult there until recently-developed technologies for direct drilling into rice stubble are proven and adopted.

Ammonia losses from lowland rice fields are estimated to account for 20 percent of all mineral N fertilizer applied, or about 2.4 Tg NH₃-N per year. The magnitude of ammonia volatilization largely depends on climatic conditions and method of application. In tropical regions, ammonia losses can reach up to 54 percent of all applied mineral N fertilizer, while emissions are generally negligible from direct-seeded rice culture in temperate regions where the majority of the fertilizer is applied prior to flooding. Volatilization of ammonia reduces the efficiency of N fertilization and increases input costs. Through aerial translocation, it can also lead to unintended N inputs into neighbouring sensitive natural ecosystems, potentially resulting in their degradation and loss of biodiversity.

Greenhouse gases

In the early 1980s, it was estimated that lowland rice fields emitted more than 50 Tg methane per year, or about 10 percent of global methane emissions. Recent measurements, however, show that many rice fields emit substantially less than those investigated in the early 1980s, especially in northern India and China. Also, methane emissions have actually decreased since the early 1980s because of changes in rice production systems such as the decrease in organic inputs. At the same time, techniques for upscaling of greenhouse gas emissions have improved greatly. Current estimates of annual methane emissions from rice fields are in the range of 10 to 30 Tg. The magnitude and pattern of methane emissions from rice fields is mainly determined by water regime and organic inputs, and to a lesser extent by soil type, weather, management of tillage, residues and fertilizers, and rice cultivar. Flooding of the soil is a prerequisite for sustained emissions of methane. Mid-season drainage, a common irrigation practice adopted in major rice growing regions of China and Japan, greatly reduces methane emissions. Similarly, rice environments with an insecure supply of water, namely rainfed rice fields, have a lower emission potential than irrigated rice. Organic manure generally enhances methane emissions.

Few accurate assessments have been made of emissions of nitrous oxide from rice fields. Emissions are primarily affected by the availability of mineral N in the soil and by soil water status, and are stimulated by the transitions between anaerobic (waterlogged) and aerobic conditions. Constant flooding generally implies relatively low emissions with the notable exception of the phase immediately after N fertilizer application. In irrigated rice fields the bulk of nitrous oxide emissions occur during fallow periods. Cumulative nitrous oxide production during the pre-rice fallow and the rice crop is 1 to 7 percent of applied N.

Crop production per se is neither considered a source nor a sink of carbon dioxide, because the carbon stored during crop growth is ultimately released after harvest through decomposition and/or burning of crop residues and through consumption of the produce. However, lowland rice culture increases the soil carbon content, and rice production can result in the net sequestration of carbon. The amount of sequestration is difficult to quantify, and the contribution of rice cultivation to the global carbon dioxide budget is unknown.

Soil and water pollution

Changes in water quality associated with lowland rice production may be positive or negative depending on the quality of the incoming water and rice cultural practices. The quality of the water leaving rice fields may be improved as a result of the capacity of the soil to hold contaminants such as heavy metals, and its ability to transform organic contaminants and trap sediments. A lowland rice field behaves as an artificial wetland in its capacity to remove nitrogen and phosphorus.

Nitrate leaching from flooded rice fields is normally negligible because of rapid denitrification under anaerobic conditions. Ammonium, the predominant form of mineral nitrogen in saturated soils, usually has a very low mobility in soil. However, high nitrogen pollution of freshwaters can be found in lowland rice growing regions where fertilizer rates are excessively high, for example in Jiangsu province in China. Nitrogen transfer from wetland rice fields to freshwaters by direct flow of dissolved nitrogen in floodwater through runoff/drainage warrants more attention.

In traditional rice systems, relatively few herbicides are used as puddling, transplanting and ponding water are effective weed control measures. In Asian rice systems in general, biocide (herbicides, pesticides, insecticides, etc.) use is small in terms of dosage and number of applications, and the chemicals degrade more rapidly in tropical flooded conditions than in temperate non-flooded conditions of non-rice crops. However, the negative human health impact of biocides is large and overwhelms the impact on the rice ecosystem and the environment. Since lowland rice fields are closely linked with the hydrologic cycle of basins, there is a high risk of discharge of biocides and their toxic metabolites to waterbodies. The potential for water pollution by biocides is greatly affected by field water management. Different water regimes result in different pests and weeds, and therefore require different amounts and types of biocides. Residual biocides interact differently with soil under different water regimes. The recent success of integrated pest management has reduced the use of biocides in some Asian countries such as Indonesia and Viet Nam. Studies on the impact of drainage waters from rice systems on downstream flora and fauna and ecosystem function are lacking.

Percolating water from lowland rice fields usually raises the water table. Where the groundwater is saline, this can salinize the rootzone of non-rice crops in the area, and cause waterlogging and salinity in lower areas in the landscape, such as in parts of the northwest Indo-Gangetic Plain. On the other hand, flooded rice can be used in combination with adequate drainage to leach accumulated salts out of the rootzone, as in parts of northern China, and to reclaim sodic soils when used in combination with gypsum as in parts of the northwest Indo-Gangetic Plain.

Farming on acid sulphate soils, which is common in the deltas of Southeast Asia, entails the risk of leaching pollutants to adjacent areas. The leaching process is especially intense under upland crops that are often grown on raised beds to avoid flooding. These leachates can adversely affect rice production through low pH and high aluminum concentrations, as observed in the Mekong Delta. Flooded rice can also be grown on acid sulphate soils, and leachates from rice fields are less harmful (having higher pH and lower aluminum levels) than those from upland crops.

Ecosystem services (multifunctionality)

The lowland rice landscape provides many ecosystem services which have been receiving increased recognition under the term “multifunctionality”. In East Asian countries/areas like Japan, Republic of Korea, Taiwan Province of China multifunctionality has become a focal issue in agricultural policy in the context of international trade negotiations. Because market forces alone are not sufficient to induce farmers to produce non-food benefits, several countries argue that they must be able to promote these beneficial outcomes without interference from the international trade bodies. However, few studies of ecosystem services of rice landscapes have been conducted to date. These few studies indicate that the multifunctional aspects vary with infrastructural, regional, climatic, social, cultural, and economic conditions. Examples of commodity functions beside the production of rice are the raising of fish in rice fields, ponds or canals, and the tending of ducks in stubble fields. Frogs and snails are collected for consumption in some countries. The non-commodity outputs, or externalities, from rice production can be both positive and negative.

Bunded rice fields may increase the water storage capacity of catchments and river basins, lower the peak flow of rivers, and increase groundwater flow. For example, in 1999 and 2000, 20 percent of the floodwater in the lower Mekong River Basin was estimated to be stored temporarily in upstream rice fields. The many irrigation canals and reservoirs associated with the lowland rice landscape have a similar buffering function. Other possible services of bunded rice fields and terraces include the prevention of land subsidence, soil erosion and landslides. Percolation from rice fields, canals and storage reservoirs recharges groundwater systems. Such recharge may also provide a means of sharing water equitably among farmers who can pump from shallow aquifers at relatively low cost rather than suffer from inequitably shared or poorly managed surface irrigation systems.

Flooded rice fields and irrigation channels form a comprehensive water network, which, together with their contiguous dry land, provides a complex mosaic of landscapes. Surveys show that rice-based ecosystems sustain a rich biodiversity, including unique as well as threatened species, and also enhance biodiversity in urban and peri-urban areas. In Asia, lowland rice fields are valued for their scenic beauty. In parts of the USA, such as California, rice fields are ponded in winter and used to provide habitat for ducks and other water birds for recreational hunting. Climate mitigation by rice fields has been recognized as a function in peri-urban areas where paddy and urban land are intermingled. This function is attributed to relatively high evapotranspiration rates resulting in reduced ambient temperature of the surrounding area in the summer, and in lateral heat emission from the waterbody in winter.

5. The challenges

Food security and poverty alleviation

The future demand for rice is a function of population growth and income-driven consumption characteristics. The population growth rate in Asia today stands at about 1 percent per year. As incomes rise, particularly in urban areas, per capita rice consumption declines. At the national level, per capita demand for rice is declining not only in East Asia (Japan, Republic of Korea, and Taiwan Province of China) but also in Malaysia and Thailand. Overall, however, the demand for rice is still growing, though at a lower rate than previously. In the mid-1990s, it was estimated that total rice production would need to increase to 750 million tonnes by 2025 and to 875 million tonnes by 2050 to meet the (changing) food demand of the growing population and to eliminate hunger among the 600 million poor in different parts of Asia. This increase in food production has to be realized with declining water, land and labour resources, and probably under increasing incidence of abiotic stresses. The challenge is to find ways to increase the productivity of scarce resources to achieve growth in total resource (factor) productivity so that future food needs can be met.

Besides quantity, the quality of rice is becoming increasingly important. More-affluent rice consumers become more discriminative with respect to taste, texture and other quality attributes. In poor rural areas, increasing the nutritional quality of rice (iron, zinc, provitamins) can help alleviate malnutrition and improve health.

Despite advances in poverty alleviation in the past few decades, absolute numbers in poverty have declined very little, especially in South Asia. Therefore, a major challenge is not only to produce more rice, but to produce it more cheaply so that the livelihoods of rice farmers can be sustained while feeding the poor. However, low rice prices threaten the livelihoods of rice farmers, the very segment of the population which helped to alleviate poverty. Therefore, costs of production need to be decreased so that both producers and consumers benefit. Lowering the cost of production has to be achieved in the face of rising wages, and the aging and feminization of rural populations. In many instances women’s role in agriculture will shift from unpaid labourers to farm managers.

Increasing productivity and protecting the environment of irrigated ecosystems

Productivity of the irrigated ecosystems needs to increase to provide sufficient cheap rice for urban and rural net food consumers. In some of the major rice-producing countries in Asia, such as Bangladesh, the Philippines, and Thailand, there is still a large gap between actual and potential yields, and efforts need to be directed at adoption of improved crop management technologies. In other countries, such as China, Japan and Republic of Korea, the yield gap is closing and efforts are needed to increase yield potential. Increased yield and increased total production mean that with current cultural practices more water will be needed to meet the increased evapotranspiration requirements. With increasing water shortage, this means that the water productivity of rice needs to increase.

Irrigated rice culture has been practised for thousands of years in various parts of Asia. This fact, together with the findings of 30 long-term continuous cropping experiments at 24 sites in Asia, suggests that irrigated lowland rice productivity is relatively sustainable, given an assured water supply. However, water shortage endangers the sustainability of lowland rice production, and may increase negative externalities. There are indications that soil-borne diseases (e.g. nematodes, root aphids) and micro-nutrient disorders are more serious threats in non-flooded than in flooded rice systems. Whereas methane emissions will decrease and nitrous oxide emissions increase under non-flooded conditions, the net greenhouse gas impact is currently unknown. Less ammonia volatilization is expected under non-flooded conditions, but more nitrogen will be present in the form of nitrate, and this nitrate may be prone to leaching or denitrification upon flood irrigation. Submergence is a key component for the relatively high carbon sequestration capacity of lowland rice fields. Although direct evidence from converted paddy fields is still missing, it is likely that growing rice with less water or conversion to non-rice crops will reduce soil carbon. This change in soil organic matter will be accompanied by changes in the microbial community, shifting from predominately anaerobic to aerobic organisms. It is not clear if or how these changes will affect soil fertility. Flooded rice has both fewer total weeds and a different weed spectrum than rice that is not permanently flooded. It is expected that water shortages will lead to increased herbicide use, as well as different types of herbicides. The challenge is to develop weed management practices that reduce the reliance on herbicides. With less water, the amounts and types of pests and diseases may change as well as predator-pest relationships. In particular, soil and root-borne diseases may increase under non-flooded soil conditions. The impacts of shifts in the use of pesticides cannot yet be predicted.

Managing water resources and irrigation infrastructures

As water becomes increasingly scarce, it must be managed as a scarce resource at farm, system, basin, and national levels. Most of the irrigation water for rice is supplied by state-operated large-scale surface irrigation systems that suffer from a legacy of poor design, degraded infrastructure, poor management, and stagnation. Currently, reformed institutions do not capture the complexities of the hydrological cycle, the multifunctionality of irrigation systems, and service relationships between different levels of management. The challenge is to transform these systems from supply-driven to demand-driven, responsive systems. Their financial, environmental, technical and service performances need to be improved to increase control, reliability, equity and flexibility. This will allow these systems to adapt to changing water allocations, and enable farmers to increase productivity, be more responsive to market opportunities, and adopt new and diversified water management practices. Improved system operation and service delivery can contribute significantly to alleviating waterlogging and salinization, but investment in drainage will be required in salt-affected areas. At the institutional level, the challenge is to develop new frameworks that can manage the complexity of the hydrological cycle, the multiple roles of irrigation systems, and deliver irrigation and drainage services to farmers in an accountable and efficient manner. At the policy level, the challenge is to harmonize water and irrigation policies with agricultural and environmental policies and integrate them into overall socio-economic development policies through strategic planning and management at national policy, river basin and irrigation system levels.

Increasing productivity and protecting the environments of rainfed and unfavourable ecosystems

The productivity of the rainfed and unfavourable ecosystems needs to increase to alleviate poverty and achieve food security for the rural poor. There is a lack of flexible crop and natural resource management techniques to cope with the complexity of the abiotic stresses in these environments. The high variability of rainfed environments exposes farmers to higher risk of losing their inputs. Furthermore, although the adoption of new varieties can be fast, the adoption of new crop and natural resource management techniques is usually slow. They often consist of different elements and farmers need time to familiarize themselves with them and become receptive to them. Although adaptive and participative research in farmers’ fields was identified as the most promising approach, clear concepts and methods are neither available nor practised in current research activities. How to transmit the targeted “conditional recommendations” to farmers is a challenge rarely addressed. Simple decision support tools for farmers could assist such site-specific management technologies, but most existing and successful tools address favourable environments. Increased flexibility will also be necessary to help farmers cope with changing and more extreme weather conditions. These could be a result of global climate change and will affect rainfed environments more strongly than irrigated systems.

Another major challenge is to avoid the potential negative environmental consequences of intensification. Agrochemicals are important ingredients of intensified systems but their use must be optimized to minimize harmful effects on the environment and human health. Possibly because rainfed systems are generally perceived as “natural”, rice growing environments where no or few agrochemicals are used and traditional varieties dominate, little research has focused on sustainability issues. Emission of methane from rainfed systems was found to be comparatively small, but nitrate accumulation in aerobic phases might contribute to considerable emission of nitrous oxide. Intensification through increased fertilizer use, cropping intensity, and changes in methods of crop establishment will affect such processes considerably. Increased productivity initially based on better varieties and subsequently on (unbalanced) inorganic fertilizer and reduced organic fertilizer use changes nutrient balances and increases the mining of soil nutrients. Reports of quickly emerging, severe nutrient deficiencies after intensification testify that rainfed systems are comparatively fragile because of frequently low natural soil fertility and low buffering capacity.

Climate change

Climate change is expected to increase the frequency of extreme events, such as storms, droughts in dry areas, and heavy rainfall and rainfall intensity in monsoon climates, which will increase the incidence of flooding in low-lying and poorly-drained areas. Sea level rises are expected to increase flood risk and salinity intrusion in delta areas. Rice ecosystems are especially prone to increased salinity intrusion and flooding because of their widespread occurrence in low-lying deltas and inlands. Despite the benefits of elevated CO₂, climate change will reduce yields in most major rice producing regions as a result of higher temperatures, and possibly because of increased UV-B radiation and increased storm damage. Recently, yield reduction in rice has been correlated with increased night-time temperatures. Higher temperatures will reduce crop water use efficiency because of lower biomass production and higher transpiration.

Valuing the ecosystem services of rice landscapes

It has become common to view the supply of environmental attributes and other non-commodity outputs as secondary factors in the pursuit of traditional policy objectives, such as food security and income support. If addressed at all, the focus is likely to be on a single environmental attribute. However, an approach that treats environmental aims and the production of non-commodity outputs as subsidiary factors is an outdated policy paradigm. Although methodologies exist to measure and estimate various services of agricultural systems, quantifying and valuing the positive and negative externalities still present a major problem. The types, magnitudes and values of the ecosystem services often vary geographically, and accurate estimates must be based on regional or local data. In many countries, relevant data at the appropriate geographic level are not available. Furthermore, there is a potential for the erroneous estimate of multifunctional outputs because of double counting, failure to recognize interactions among the outputs, and failure to consider the potential outputs from other uses of the land.

6. Response options

6.1 Varietal improvement

The complete sequencing of the rice genome is expected to accelerate the discovery and exploitation of useful genes in breeding programmes for all ecosystems. However, to enhance identification and adoption of superior varieties, farmers’ needs, preferences, and opinions should be taken into account in the selection process, e.g. via farmer-participatory variety selection.

Varietal improvements in irrigated systems

The key factors in the success of the green revolution varieties were increased harvest index and short duration while maintaining yield potential. However, there are no indications that these factors can be further exploited to significantly increase the yield potential of inbred varieties under fully irrigated conditions in the future. For example, since the introduction of IR8, the yield potential of semi-dwarf tropical indica inbred rice varieties has stagnated at about 10 t ha^-1. The yield potential of longer-duration temperate japonica varieties is around 14 t ha^-1. Significant yield improvement has recently come mainly from the development of hybrid rice, which has increased yield potential by 5 to 15 percent over inbred varieties in the same ecosystem. Transforming the C3 rice plant into a C4 plant by genetic engineering could be a long-term approach for increasing rice yield potential.

The reduced growth duration of modern high-yielding rice varieties has reduced total outflows of evaporation, seepage and percolation from rice crops. The combined effect is that these varieties have a water productivity with respect to total inputs (irrigation, rainfall) that is three times that of traditional varieties grown under a similar water management regime. The higher harvest index contributed to increased water productivity through higher grain yield per unit water transpired or applied. As for yield potential, there appears to be little scope for further increasing water productivity by further reduction in growth duration or increased harvest index under fully irrigated conditions.

Traditional breeding programmes for irrigated environments have been selected under conditions of continuously ponded water. With increasing water scarcity in irrigated systems, breeding programmes should also include selection under conditions of water saving technologies such as alternate wetting and drying or aerobic cultivation (see below). Some success has been recorded with the development of high-yielding aerobic rice varieties in northern China. Breeding programmes for water-short conditions should focus on maintaining harvest index, seedling survival, and early vigour. A variety of breeding strategies can be explored to increase water productivity with respect to evapotranspiration, such as early vigour to reduce soil evaporation, weed suppression to reduce weed transpiration, and increased waxiness of leaves to reduce non-stomatal transpiration. The modern, improved japonica varieties have higher transpiration use efficiencies than the older indica varieties, suggesting that significant variation exists for this trait in rice germplasm. The potential for exploiting this trait, however, has not been investigated. Transforming rice into a C4 plant could potentially also increase water productivity, though no evaluation of the effectiveness of such an approach has been made.

Varietal improvements for tolerance to abiotic stresses

For drought conditions, most progress so far has come from the development of short-duration varieties that escape drought at the end of the rainy season. But recently, substantial genetic variability for grain yield under drought stress has been documented and demonstrated to be a moderately heritable trait, with repeatabilities similar to those of yield in non-stress environments. New breeding approaches and improved screening methods are advancing the development of drought-tolerant varieties and need to be expanded to national programmes.

Though breeding for submergence tolerance and enhanced yield in flash-flood areas has been going on for over three decades, only a few tolerant lines with improved agronomic characteristics have been developed so far. However, fast progress is being made with the development of submergence tolerant lines using marker-assisted selection. A first major quantitative trait loci (QTL) was fine-mapped, and markers were developed and successfully used to transfer this QTL into a popular rainfed lowland variety. Efforts are currently ongoing to transfer this QTL into other lowland rice varieties. For deepwater areas, some breeding progress has been made and a few new lines with reasonable yield and grain quality have been released. Recently, three main QTLs for elongation ability were identified. Fine-mapping and tagging of these QTLs should facilitate their efficient incorporation into modern popular varieties through marker-assisted selection.

Despite its high sensitivity to salinity, considerable variation in tolerance exists in rice. Combining new efficient screening techniques with conventional, mutation and anther culture techniques, salinity tolerance was successfully introduced into high-yielding plant types. Some newly released varieties have demonstrated more than 50 percent yield advantage over current salt-sensitive varieties. Breeding cultivars with much higher tolerance is possible if component traits are combined in a suitable genetic background. The opportunity to improve salinity tolerance through the incorporation of useful genes and/or pyramiding of superior alleles, appear very promising. A major QTL was recently mapped and marker-assisted backcrossing is currently being used to incorporate this QTL into popular varieties that are sensitive to salt stress.

Building on this, improved modern-type varieties with an increase in yield potential of about 1 t ha^-1 should become increasingly available and widespread in drought-, flood- and salinity-prone rice areas within the next ten years. A careful characterization of target areas will be essential for developing effective varieties and management technologies.

Varietal improvement for tolerance to climate change

Preliminary results suggest that there is genotypic variation in the sensitivity to warm night temperature. If this is true, there is a possibility to develop rice varieties that are insensitive to warm night temperature through plant breeding. Moderately large genetic variation in the tolerance to high-temperature-induced spikelet sterility exists among rice genotypes. Selection of rice varieties that flower early in the morning can be an effective way to avoid high-temperature-induced spikelet sterility. Therefore, both tolerance and avoidance mechanisms can be used in breeding varieties that yield well under extremely high temperature. Significant intraspecific variation in yield response with a doubling of current CO₂ levels has been observed in rice. The CO₂ responsive characteristics of older cultivars could, potentially, be incorporated as factors in future varietal selection to maximize the beneficial effect of carbon fertilization.

6.2 Quality improvement

A growing trend in breeding is the effort to increase not only crop yield potential, but also the nutritional value of the food derived from the crop. To alleviate malnutrition, a particular emphasis is to increase the micronutrient density of the edible portions, with specific focus on increasing provitamin A carotenoid, iron, and zinc contents. It still remains to be seen, however, whether these contents can actually be increased in the endosperm as opposed to the bran, to make a significant impact on human health and nutrition, given the almost universal compulsion to eat polished rice. The efforts to increase nutritional value are not expected to have any effect on the water relations of the crop. To drive the adoption of these new varieties, the successful traits will need to be combined with other successful traits, including those that impart drought tolerance. This is especially pertinent as the major target farmers and consumers for more nutritious rice are located in the most disadvantaged zones (rainfed and unfavourable environments), both in terms of poverty and crop production resources.

7. Field management to increase yield and water productivity

Water-saving technologies and holistic approaches for irrigated ecosystems

A suite of so-called “water-saving technologies” exists, or is being developed, to assist farmers to cope with various degrees of water scarcity in irrigated ecosystems. These water saving technologies increase water productivity with respect to total water inputs (rainfall, irrigation), mainly by reducing seepage and percolation flows, and to a lesser extent by reducing evaporation. General measures such as land leveling, farm channels, good puddling and bund maintenance improve water control and reduce seepage and percolation outflows. Minimizing the turnaround time between wet land preparation and transplanting can be accomplished by the adoption of community seed beds or adoption of direct seeding. Direct dry seeding can increase the effective use of rainfall and reduce irrigation needs. Mechanical soil compaction can reduce percolation flows in certain soil types. Chemical evaporation suppressants have been tested in flooded rice fields but financial and environmental impacts have not been assessed. Water management techniques such as saturated soil culture and alternate wetting and drying, can reduce field water application by 15 to 20 percent without significant impact on yield. Alternate wetting and drying has been widely adopted throughout China and can now be considered the common practice of lowland rice production. In the system of aerobic rice, specially adapted “aerobic rice varieties” are grown under dryland conditions just like other cereals such as wheat, with or without supplemental irrigation. Currently, yields of aerobic rice systems can be up to 20 to 30 percent lower than continuously flooded systems with conventional lowland varieties, while water inputs can be reduced by 30 to 50 percent. In aerobic rice systems, resource-conserving technologies, as practised in upland non rice crops, become available to rice farmers as well, such as mulching, zero tillage and minimum tillage. Growing rice under aerobic conditions on raised beds shows promise but is still in its infancy with many challenges to be overcome, including micronutrient deficiencies, biotic stresses, yield sustainability and identification of optimum irrigation management where limited irrigation water is available. The development of management technologies needs to be integrated with characterization of the water-short target environments and the selection or breeding of appropriate varieties for target environments. The example of aerobic rice is a case in point.

Increasingly, technologies that aim to close the gap between actual (farmers’) yields and yield potential need to apply holistic approaches that integrate various components of crop, soil and water management. Other examples are the system of rice intensification (SRI), site-specific nutrient management and integrated crop management. Depending on baseline conditions, yield increases of 10 to 100 percent have been reported with these technologies. Though water flows have hardly been studied in these integrated yield-improving technologies, it is likely that yield increases are accompanied by relative increases in transpiration, and by relative decreases in evaporation, seepage and percolation. In terms of water savings, any agronomic practices that increase harvest index will result in more grains per unit water transpired and hence increased crop water productivity. The system of rice intensification includes the water saving practice of alternate wetting and drying and has been shown to decrease water inputs at the field level. The appropriateness of each (integrated) technology or system depends on the nature and severity of water scarcity, farmers’ current practices, soil properties, and hydrological boundary conditions (groundwater table depth, rainfall). The attractiveness of the technologies depends on profitability, risk perception and ease of adoption by farmers. The technologies need to be carefully targeted and widely tested and improved using farmer participatory approaches.

Sustainability and environmental protection under water scarcity

While relatively much work has been done on the development of technologies to increase crop productivity under water scarcity, little attention has been paid to long-term sustainability and to the reduction of negative environmental impacts of rice production systems that use less water. Studies are needed on the relationships between type, amount, timing and method of application of organic and inorganic N fertilizers, water management and crop residue management on the one hand, and yields and yield sustainability, greenhouse gas emissions and pathways of N losses on the other. The effectiveness of fertilizer management technologies such as site-specific nutrient management, leaf color chart, slow-release fertilizers, and deep placement, need to be evaluated under various conditions of water scarcity from saturated soil to alternate wetting and drying to fully aerobic soil conditions. Little is known about changing pest dynamics when field conditions change from water-abundant conditions to water-short conditions. Under increasing aerobic soil conditions, pest and disease management technologies used for dryland crops may become relevant for rice. Technologies based on integrated pest management need to be tested under water-short conditions. The development of appropriate crop rotations needs attention.

Increasing productivity in the rainfed and unfavourable ecosystems

The development of varieties that are tolerant to abiotic stresses will have considerable impact on many unfavourable environments. Input-responsive varieties, shorter duration, and reduced risk increase the incentive to use external inputs and to intensify the cropping system. Adjusted management technologies will be needed to make the best and most efficient use of the possibilities offered by the new varieties. These technologies should aim at reducing stress intensity, enhancing survival and robustness of the crop to withstand stress and stabilize yields, and avoiding stress occurrence at sensitive crop stages. Because of the complexity of the systems to be targeted, very close cooperation between breeders, agronomists, social scientists, and farmers will be necessary to produce relevant results.

For the drought-prone rainfed lowlands, two promising technologies are direct seeding and improved nutrient management. Direct seeding potentially offers better use of early-season rainfall, better drought tolerance as a result of better root development, lower risk from late season droughts, better use of indigenous soil N supplies, and an increased chance for a second crop after rice. Site- and season-specific nutrient recommendations can reduce nutrient losses and chemical pollution of the environment by allowing for the natural soil fertility and the likely field water supply, combined with an appropriate rice variety. Both technologies have already enabled substantial productivity increases in some more favourable rainfed systems. In Indonesia (Lombok), the introduction of short-duration and input-responsive varieties with direct seeding and the use of inorganic fertilizer increased and stabilized yields in the gogorancah system. In Lao PDR, rainfed lowlands contributed considerably to achieving self-sufficiency in rice within a decade after the introduction of improved varieties and crop management. Similar successes should be feasible in drought-prone lowlands in eastern India. Smaller and slower progress characterizes more unfavourable environments, but while yields and cropping intensity have been almost stagnant in most irrigated systems for the past two decades, they started to increase faster in rainfed systems about twenty years ago. Technologies based on integrated pest management should also be further explored for the drought-prone lowlands.

In the submergence-prone environments, the combination of new submergence tolerant varieties with adapted crop management and nutrient management can improve seedling and plant survival as well as the ability to recover after submergence. Seedlings enriched in nutrients, particularly zinc and phosphorus, and possibly silicon, have greater chances of survival because they have enhanced growth and accumulated higher carbohydrate reserves. Application of certain nutrients after the recession of flood water also helps in faster recovery, better tillering and higher grain yield.

Because the ponded water leaches accumulated salts from the topsoil, lowland rice is the only cereal cropping system that has been recommended as a desalinization crop. Soil amendments, particularly gypsum, assist in reclaiming salt-affected soils but require large investment. New approaches involving the use of farmyard manure, pressmud from industrial waste, and tolerant varieties can help reduce the need for gypsum by more than 50 percent. The integration of tolerant varieties with specific nursery management, crop and nutrient management strategies are needed to enhance crop survival and productivity, mitigate salt stress and improve soil quality.

Lowering cost of production and labor requirements

Many integrated crop management practices aim not only to increase yield, but also to increase the efficiency of resource inputs, thereby lowering the cost of production. Where irrigation water is supplied by pumping, water-saving technologies reduce water inputs, pumping costs and energy consumption. Whether this actually increases profitability depends on the yield obtained and the relative price of rice and water/pumping. Reducing the frequency of irrigation reduces the labour use for irrigation. On the other had, when fields are not continuously flooded, weed infestation may increase and more labour or herbicides may be required. The system of rice intensification has relatively high labour requirements and abandonment of the system has been reported in its country of origin, Madagascar. Conversely, the adoption of labour-saving practices can impact on the way in which water is used. For example, studies in the Philippines and Malaysia show substantial irrigation water savings with dry (direct) seeding. However, it is not clear whether changing to direct seeding will save water everywhere. Dry seeding and aerobic rice technologies offer possibilities for mechanization of farm operations such as seeding, weed control and combine-harvesting (which is easier on firm dry land than on lowland soil that may still be relatively wet). So far, however, few studies have reported the trade-offs between reduced inputs (e.g. water, labour, energy) and impacts on yield, and more effort is needed to identify win-win (e.g. more profitable, more convenient to the farmer) situations. With increasing feminization of the rural labour force, women need to be involved in the development of suitable production technologies.

8. Options at the landscape level

Irrigation systems

Rice irrigation systems are characterized by large volumes of water in circulation through surface drainage, seepage and percolation from the flooded rice fields. In many systems in low-lying deltas or flood plains with impeded drainage, the continuous percolation of water has elevated groundwater tables close to the soil surface. Water-saving technologies at the field level mainly reduce seepage and percolation flows from the field. Because these flows can often be recaptured downstream, they may not contribute to water savings at higher spatial scales such as the irrigation system. Recent studies of rice-based irrigation systems in China, the Philippines and Sri Lanka indicated that irrigation efficiency improves with increased area of spatial domain because of the reuse of water. In some situations, reducing percolation from rice fields can lower groundwater tables and negatively affect yield-water application relations at the field level, while increasing the cost of pumping for reuse downstream. On the other hand, lowering of the groundwater table may reduce direct evaporation from the groundwater, which is a true water-saving since it reduces a non-beneficial depletion flow. Evaporation may also be reduced from rice fields using water-saving technologies. Alternate wetting and drying has been found to reduce evaporation by 0 to 30 percent, compared with reductions of 50 to 75 percent in aerobic rice. Overall, there is still very little understanding of the potential impacts of adopting field-level water-saving technologies on the water dynamics, water balance and water productivity of irrigation systems as a whole.

More systematic evaluations are needed to understand the water balance dynamics and productivity of water at various scales of the irrigation system. Application of new concepts of irrigation efficiency that take into account the return flows (irrigation water run-off and percolation that re-enters the water supply to fields), are needed in designing interventions for improving total water use efficiency at the system scale. It is important to select the appropriate set of water balance and water productivity indicators, specify the spatial domain of interest and the interactions (agrohydrological, socio-economic and environmental) across spatial and temporal scales among the users of water. A system’s approach is essential to determine water balance related objectives and water management strategies to achieve them. These strategies should aim at improving water control, equity, reliability and flexibility of service to allow farmers to make choices of crop and water management. Major options for improving irrigation systems include conjunctive use of surface and groundwater, recirculation of drainage, buffer reservoirs at appropriate levels in the systems, improved design of control structures, investment in drainage, operation and ordering procedures and intensification of irrigation system management. The gap is in capacity building of the irrigation profession at large and a critical required action is the revision of design standards.

Rainfed and unfavourable environments

There are many interventions at the landscape level that are effective in alleviating field-level abiotic stresses, and in increasing land and water productivity in rainfed and unfavourable environments. On-farm water harvesting is an effective means of reducing risk and increasing productivity in drought-prone rainfed environments. The development of reservoirs and canal networks to store fresh water from rain or rivers before they become saline can help extend the growing season in saline coastal areas and can substantially improve productivity. The main constraints to technology adoption are related to socio-economic and organizational issues. Therefore, increased use of this option seems to be more dependent on local decision-makers and national governments than on any other factor.

Water management through large-scale construction of coastal embankments and sluices has been reasonably successful in preventing seawater intrusion in many deltaic coastal areas, substantially reducing soil salinity in the wet season. Coastal embankments have increased rice yields in coastal Bangladesh by more than 200 percent in 20 years. The technology also opens up the possibility of growing high yielding, modern rice varieties in these areas, as in the coastal areas of the Mekong Delta, Viet Nam. However, water needs to be managed judiciously to avert undesirable long-term environmental consequences and local conflicts with other water users, especially the landless poor who depend on brackish water fisheries for their livelihoods.

Ecosystem services (multifunctionality)

The concept of ecosystem services, or multifunctionality, was conceived to give recognition to the environmental attributes of crop production such as rice production. More recently the interest in multifunctionality has taken on a policy emphasis as governments, particularly in East Asia, seek to provide a rationale for farm subsidies. However, as we move to an era of scarcity and increasing competition for water, greater emphasis needs to be given to both positive and negative externalities of water in rice-based landscapes. This will require greater emphasis on appropriate data collection to facilitate the estimation of ecosystem services. More case studies are needed to identify local or regional specific characteristics of ecosystem services, and management practices need to be developed to sustain and enhance the ecosystem services of rice-based landscapes. Finally, agricultural policies should be implemented that explicitly consider the multiple outputs of rice production.

9. Creating incentives for improved water management

Farm and village level

For farmers who are already directly confronted by water scarcity, further incentives to improve water management may not be necessary — the cost of acquiring water is sufficient incentive. For example, many farmers using pump irrigation already face a positive marginal cost every time they irrigate their fields. On the other hand, where gravity flow irrigation is provided free to farmers, or at low fixed fees, or where electricity for pumping is subsidized, farmers lack incentives to manage water carefully. This is not a problem where water is locally abundant. However, in many cases, water is abundant upstream but scarce downstream (for either farm uses or non-farm uses) and the lack of incentives to save water upstream becomes a problem. One option is to volumetrically price water. However, the majority of the world’s rice is grown on small farms and the cost of volumetrically pricing at the farm level is prohibitive. In this situation, it is possible to volumetrically price to the village or irrigation association. The savings may be retained by the village or association or rationed among users.

Allocation among sectors

Where there is sufficient demand for water from the industrial and domestic sectors, water is often allocated out of agriculture by administrative fiat. Administered prices or tradable water rights may provide an incentive for reallocation of water at the irrigation system or basin level. The incentive for irrigation managers to reallocate water can impose scarcity at the farm level and encourages farmers to adopt water-saving practices, technologies, or cropping systems.

Virtual water

Other solutions to water scarcity skirt the difficult problems of raising prices, instituting a system of legal rights, or administratively restricting water allocations. One of the most promising is trade in “virtual water”. This involves trading commodities where the production is relatively water intensive (e.g. rice) instead of trading water itself for which transportation costs are usually prohibitive. In the world rice economy this already happens to an important extent. The water-scarce Middle East is one of the biggest rice importing regions of the world. Mainland Southeast Asia (Myanmar, south China, Thailand and Viet Nam) has an abundance of water resources, and rice exports from these countries can help alleviate water shortages in other countries. But international trade is still restricted, and much more trade in virtual water could profitably occur.

Public and private sector

There is a growing interest in irrigation management transfer, and governments are attempting to transfer management and some of the financial burden for operation and maintenance of irrigation systems to water user groups. Meanwhile there has been a rapid private sector increase in pumps and tubewells, much of it within surface irrigation systems. There is a need to integrate the management of surface and groundwater resources to avoid overexploitation, equitable access to water, and to ensure that users have the flexibility to shift from rice to high-value crops. Given the financial constraints facing the public sector in many economies, it may well be that in the future the private sector will play a stronger role in operation and maintenance, and water users will pay private sector entities for services delivered.

Rural development

We have emphasized the challenges caused by urbanization: the ageing and feminization of the agricultural sector. Nevertheless, close to half of the labour force and the majority of the poor remain in rural areas. Fostering rural development requires investment in rural infrastructure (roads, electricity, communications, and irrigation) and in human capital (education and health). Such investments will result in employment and income gains in the rural non-agricultural sector strengthening the demand for both agricultural and industrial products.

New priorities for research

With the decline in funding for research, those areas with potential for increasing production must be carefully targeted. Scientists disagree on the potential for increasing production in the irrigated as opposed to the rainfed areas. However, the need for production increase is clear in both the irrigated and rainfed areas. Gains in production in the past have typically come from increasing area and land productivity (yield). However, as other resources have become scarce, productivity must be examined in a broader context. Reflecting the growing scarcity of labour and water, development and dissemination of labour-saving and water-saving management practices and drought-tolerant rice varieties are priority areas for research.

David Dawe³

1. Introduction

In a market economy, the most important function of agriculture is to provide food for people. Consumer demand is thus an important driver of agriculture. Tastes and preferences obviously vary from country to country and even between provinces or states within a country. Nevertheless, incomes permitting, consumers the world over have shown a nearly universal desire for a varied diet. If farmers cannot produce the variety that private consumers want at reasonable prices, then agriculture is doomed to stagnation, and it will ultimately be unable to provide a sustainable living to farmers. Of course, it is possible for farmers to produce what the government wants (instead of what consumers want) for some period of time, but government demand and subsidies are not likely to be sustainable. For example, after many years of farm subsidies, the New Zealand Government was eventually forced to abandon them in 1984 in order to avoid a fiscal crisis.

Whereas consumers want to diversify their diets, the vast majority of farmers want to leave agriculture altogether whenever it is possible. They want to leave agriculture in order to escape the drudgery of farm labour and to have a chance at earning the higher incomes that are possible in non-agricultural work. There are not many farmers in Asia whose ambition it is to have their children continue to work in the fields. Farmers understand that in the economy of the future, it will not be enough simply to own one or two hectares of land. In other words, if land is the only asset one possesses in addition to the physical power of the body, it will be impossible to escape poverty.

Economic growth is the most important factor that determines whether or not consumers and farmers can satisfy these desires. Fortunately, despite the financial crisis that gripped the region in 1997, economic growth has been quite rapid in most of Southeast Asia during the past 40 years or more (Figure 1). These rates of growth cumulate strongly over long periods of time. For example, per capita GDP in Thailand, after adjusting for inflation, was nearly seven times higher in 2003 than it was in 1960.

In addition to rapid income growth, urbanization is also proceeding rapidly (Table 1). It appears to be proceeding more rapidly in insular Southeast Asia (Myanmar, Indonesia, Malaysia, and the Philippines) than in mainland Southeast Asia (Cambodia, Lao PDR, Myanmar, Thailand, and Viet Nam). As urbanization proceeds and job opportunities outside of agriculture expand, farms will need to become more commercially oriented in order to supply food to people who no longer work or live on farms. As consumer demand for convenience increases, agro-industries will develop to create more highly processed food, and farmers will need to pay more attention to quality in order to sell their output.

In addition to affecting consumer demand for food, and farmers’ occupational choices, economic growth and urbanization also profoundly affect the supply of water available for agricultural use, because of rapid growth in demand for water for industrial, household, and environmental purposes. The objective of this paper is to explore what considerations agricultural water managers will need to take into account in order to effectively deliver services to farmers and society, given the rapidly changing world around them that is being created by economic growth and urbanization. The first section will explore the nature of three key macro trends — the declining relative importance of agriculture, changes in consumer food demand, and increased trade. The second section will discuss the implications of these macro trends for farm households and irrigation system managers.

Notes: Because of lack of data availability, growth rates were calculated for shorter periods as follows: Cambodia (1993 to 2003), Lao PDR and Viet Nam (1984 to 2003), Myanmar (1960 to 2001).
Source of raw data: World Bank (2005).

Figure 1. Average annual growth rates of per capita GDP, 1960 to 2003

Table 1. Urbanization rates in Southeast Asia, 1961, 1990, and 2004

Source of raw data: FAO (2005).

2. Key macro trends driving Southeast Asian agriculture

2.1 The declining relative importance of agriculture

Along with the economic growth noted above, the importance of agriculture to national economies, measured as a percentage of GDP, has consistently declined (Figures 2a and 2b). The share of agriculture in the labour force, although greater than the share of agriculture in GDP, is also declining (Figures 3a and 3b). This decline in the relative importance of agriculture is not unique to Southeast Asia — it is a fundamental feature of what is known as the structural transformation in economic development, and it has occurred in all countries around the world that have experienced economic growth.

This structural transformation does not mean that the agricultural sector has contracted. Indeed, the agricultural sector has grown substantially in Southeast Asia since at least as far back as 1960. For example, the value of agricultural output in Thailand is nearly five times as large today as it was in 1960. In both Indonesia and the Philippines, it is more than three times as large. However, its growth has been less rapid than that of the industrial and service sectors, so its relative importance has declined.

Source of raw data: World Bank (2005).

Figure 2a. Agriculture’s share in GDP, 1960 to 2003

Source of raw data: World Bank (2005).

Figure 2b. Agriculture’s share in GDP, 1960 to 2003

The decline in relative importance suggests that agriculture will receive a diminishing share of government resources, as indeed it should. This may or may not mean a decline in absolute levels of funding, because economic growth can make the pie bigger for everyone. But regardless of whether government support to agriculture increases, declines or remains roughly constant, agriculture remains a key sector of the economy. It still comprises anywhere from 10 percent of the economy (in Malaysia and Thailand) to nearly 50 percent in Lao PDR. Given this importance, continued agricultural growth is critical for the overall health of the economy and for poverty alleviation. But it will be a challenge to maintain this growth and services to farmers in the face of shifting budget priorities.

Source of raw data: FAO (2005).

Figure 3a. Agriculture’s share in the labour force, 1961 to 2003

Source of raw data: FAO (2005).

Figure 3b. Agriculture’s share in the labour force, 1961 to 2003

2.2 Consumer food demand is changing

Income growth leads to many changes in diets. For example, Bennett’s Law states that the proportion of starchy staple foods (grains, roots and tubers) in total calories declines as income increases (Timmer et al., 1983). This decline has occurred in all countries in Southeast Asia during the past twenty years (Table 2; FAO, 2004).

Table 2. Percentage of calories coming from starchy staples, 1979–82 and 2000–02

Source of raw data: FAO (2005).

Within the category of starchy staples, diversification also occurs with increases in income. For example, in southern India, the importance of rice is declining as consumption shifts toward wheat. But, in northern India, where wheat is the traditional staple, demand is shifting from wheat to rice (Pingali and Khwaja, 2004). A similar trend is taking place in northern and southern China. In Southeast Asia, where rice is the staple food and wheat is not even grown, a shift from rice to wheat has been taking place in several countries. The process is well underway in Indonesia, Malaysia, the Philippines and Thailand, is just starting in Viet Nam, but has not yet begun in Cambodia, Lao PDR and Myanmar (Figure 4). Some of the shift from rice to wheat products may also be a result of increased convenience in terms of less preparation time, which is important as populations become more urbanized.

Figure 4. Calories from wheat relative to calories from rice (percent), lagged three year moving average, 1963 to 2003

The shift toward wheat and away from rice in Southeast Asia is because of demand shifts for both rice and wheat. First, per capita demand for rice has stagnated in most countries, and has sharply declined in Malaysia and Thailand, the two wealthiest economies in the region. The only factor now sustaining growth in rice demand in Southeast Asia is population growth, but this is slowing in most of the region. Second, per capita consumption of wheat has increased strongly since the middle of the 1980s, when world market wheat prices began to fall to very low levels. Ignoring the years of the world food crisis (when prices were very high), world wheat prices averaged approximately US$367 per tonne (in 2004 dollars) from 1957 to 1981. Just four years later, by 1985, prices had declined to US$238 per tonne, and they remained in the vicinity of US$200 per tonne for most of the next decade, before they fell again to current levels of about US$150 per tonne. This decline in prices made wheat much more affordable to the people of Southeast Asia, accelerating a process that would have occurred anyway, eventually. For example, Malaysian wheat consumption has been on an increasing trend for at least the past forty years, well before the fall in wheat prices or the beginning of globalization. Indeed, the increase in wheat consumption during the past two decades occurred even while the world price of wheat increased relative to the world price for rice by a substantial margin (i.e. world rice prices declined even faster than world wheat prices).

Despite the shift away from rice and towards wheat, rice will remain the staple food for the foreseeable future in Southeast Asia. Even in Japan, where income levels are much higher than in Southeast Asia, rice still provides considerably more calories than wheat. Nevertheless, rice will not be a source of dynamic consumer demand in the region, and wheat cannot be grown profitably in Southeast Asia. Thus, farmers will need to explore alternative crops to meet consumer demand.

During the past twenty years, there has been a sharp increase in consumption of fats also. For Southeast Asia as a whole, average daily fat consumption increased from 29 grams in 1975 to 54 grams by 2002, with all countries experiencing increases. Much of the increased consumption has come from animal products, but a large proportion also has come from vegetable oils (animal products supplied 30 percent of fat in 2002, compared to 28 percent in 1975). The percentage of calories in the diet coming from fat also has increased substantially since 1980 in most countries in the region, although levels remain well below the 37 percent observed in the United States. It is an open question how far this trend will continue. For example, it reached a peak of 33 percent in Malaysia in 1990, but has declined since then to a level of 26 percent. In Japan, the percentage is 28 percent, much below levels in Australia (39 percent) and New Zealand (32 percent). Thus, it is not clear that Asian diets will become as fatty as those in Western countries. However, it does seem likely that fat intake will continue to rise in Southeast Asia for some time to come, as the current percentage of calories coming from fat in the region is just 18 percent.

Protein consumption also increased substantially from 1975 to 2002, from 46 to 63 grams per capita per day. Again, all countries in the region experienced increases. The percentage of protein that comes from animal sources increased from 23 percent to 27 percent since 1986.

2.3 Increasing domestic and international trade

Whereas the agricultural sector in general will need to become more diversified to meet demand, individual farms and regions may become increasingly specialized in particular crops. Trade, both domestic and international, will thus play a greater role in supplying consumers. Indeed, the importance of trade has increased rapidly in Southeast Asia during the past 20 years. International trade in agriculture (the average of agricultural imports and exports) increased from 47 percent of agricultural GDP in 1981–83 to 89 percent in 2001–2003.

Although data on domestic trade flows are not as readily available, the available literature shows that domestic markets have become more integrated. This is not surprising given the massive increases in domestic communications and transportation infrastructure (Rashid et al., 2005). Even for an export-oriented crop like oil palm in Indonesia, nearly one-third of the massive increase in production since 1970 has been used for domestic supply, which usually involves transportation from producing areas, e.g. Sumatra, to consumption areas on Java.

3. Implications for farm households and irrigation system managers

3.1 Crop diversification and competition for water

Within agriculture, staple crops are becoming less important because of changing demand and downward trends in rice prices on the world market. For example, Isvilanonda et al. (2000) conducted repeat surveys of three villages in the province of Suphan Buri in the Central Plain of Thailand in 1987 and 1998. They report that the percentage of farmers that exclusively planted non-rice crops increased from 1 percent in 1987 to 17 percent in 1998. A similar, but smaller, change occurred in Khon Kaen Province in the Northeast. For Southeast Asia as a whole, the share of agricultural cropped area planted to starchy staples (rice, coarse grains, roots and tubers) declined 15 percentage points (from 70.6 percent of total area to 55.5 percent) between 1961 and 2004. Thus, farmers will increasingly demand irrigation that is more flexible (e.g. suitable to crops other than rice and cereals).

For the many farmers who continue to plant rice (rice still accounts for 43 percent of total cropped area in Southeast Asia), more reliable irrigation may be needed to prevent plant stress in the face of increased competition for water use from other sectors. On average, competition from other sectors may not be a major problem in Southeast Asia. Actual consumptive use of water for irrigation is expected to grow more slowly than for other uses, but it is still expected to grow, not shrink (from 85.5 km³ in 1995 to 91.9 km³ in 2025; see Table 4.6 in Rosegrant et al., 2002). Nevertheless, in certain circumstances, increased competition for water from industry and households will be a constraint for farmers. This competition will be particularly acute near urban areas and in years of drought (caused, for example, by El Niño).

Farmer demand for increased flexibility and reliability in water deliveries has contributed to the spread of pump irrigation, as has the increased availability and lower prices of such pumps, many of which come from two concentrated industry clusters in China, one in Zhejiang and the other in Fujian (Huang, 2004). In Viet Nam, the number of pumps more than quadrupled in just 11 years (1988 to 1999), from 124 thousand to nearly 800 thousand. Pumps are especially common in southern Viet Nam (Viet Nam General Statistical Office, 2000). Similar trends have been documented for Bangladesh. In the Philippines, approximately 23 percent of rice farms now use pumps to access water, either from subsoil reservoirs, drainage canals, or natural creeks and rivers. Among these three sources, groundwater is the most important (Dawe, 2005). The increased prevalence of pump irrigation is creating challenges for surface water managers to manage water conjunctively.

3.2 Labour scarcity and the increasing importance of non-farm income

In parallel with trends in the macro economy, farm households throughout Asia are diversifying their income sources outside of agriculture. Hossain et al. (2000a) report that the percentage of household income coming from agriculture in Central Luzon, Philippines, declined from 64 percent in 1985 to 40 percent in 1997. In Bangladesh, the share of rural household income coming from agriculture declined from 63 percent in 1987/88 to 54 percent in 1994/95 (Hossain et al., 2000b).⁴

Labour scarcity is increasing in most of Asia, making it more difficult to hire farm labour. For example, agricultural wages in the Philippines, adjusted for inflation, were 60 percent higher in 2002 than in 1981. Fan et al. (1999) show data for India that rural real wages were 63 percent higher in 1993 than in 1970. Sombilla and Hossain (1999) show similar trends for Thailand and Bangladesh. Thus, although demands for flexible and reliable irrigation increase, farmers will be less willing to devote their own scarce time to manage irrigation (because of the increased importance of non-agricultural income in their livelihoods noted above) and will find it difficult to hire labourers to do it for them. Increased labour scarcity and the increased importance of non-agricultural income to rural livelihoods have deep implications for the design of schemes (e.g. participatory irrigation management and irrigation management transfer) to devolve managerial and financial responsibility for irrigation systems. Although governments have fiscal incentives to devolve such management, it is not clear that rural households have incentives to spend the time required to manage these systems effectively. This problem may account for the fact that studies of participatory irrigation management (PIM) and irrigation management transfer (IMT) have typically been unable to demonstrate gains in efficiency as a result of the new management system (e.g. Vermillion, 1997).

An alternative institutional arrangement that shows some promise is described and assessed in Wang et al. (2003). In some irrigation systems in China, surface water management for a village (or other unit) is contracted to a single individual, the water manager. Farmers in the village pay water fees based on historical averages of how much water they have used, and these fees are given to the water manager. This system gives no incentives for farmers to conserve water. On the other hand, the irrigation manager is required to pay only for the water delivered to the village. If the water delivered to the village is below historical norms, then the water manager can keep the difference, generating incentives for the water manager to conserve water without reducing deliveries so much that farmers demand his replacement. Using multivariate statistical analysis, the authors found that systems where water managers have these types of incentives use less water, after controlling for a variety of other influences. This is an emerging institutional form in China, and it is not clear how widely it has spread. It is very similar to a system of water rights (Rosegrant and Binswanger, 1994), but instead of vesting such rights legally in each farmer (which might be very expensive when farm sizes are small, the rights are vested in the community, which can then temporarily assign them to individuals. Such institutional innovation has the potential to be very effective at conserving water (Dawe, 2005).

3.3 Land consolidation, aging, and feminization

Increased commercialization and specialization, along with migration to urban areas, is likely to result in land consolidation and larger farm sizes in the future. Such consolidation might make it easier to manage irrigation water deliveries, but it should be noted that the process is proceeding slowly. For example, even in Japan, a relatively wealthy country where one might expect the process to be most advanced, average farm size has increased from 0.99 hectares in 1956 to 1.59 hectares in 2003 (raw data from MAFF, 2004). This represents a cumulative increase of about 60 percent, which seems large, but in absolute terms the increase of 0.6 hectares in nearly half a century seems quite small. The situation is similar in the Republic of Korea, where average farm sizes have been steadily increasing from 1970 (0.88 hectares) until 2002 (1.46 hectares; Fan and Chan-Kang, 2003).

Land consolidation is already underway in some parts of Southeast Asia. In the Muda Irrigation System in Malaysia, many groups of farmers have contracted out operations of their farms to others, who then manage the farms as a block. Here, although there has not been large scale transfer of ownership because of land laws, management consolidation is being achieved.

Farm sizes in Thailand at the national level are still declining, according to the most recent data from the 2003 Agricultural Census.⁵ But in some dynamic agricultural areas of the country, such as Suphan Buri Province in the Central Plain about 150 kilometres north of Bangkok, average farm sizes increased from 3.8 hectares in 1993 to 4.0 hectares in 2003. This is not a large increase, but it is an increase nonetheless, and it is likely the beginning of further consolidation. The total number of farm holdings declined by 11.5 percent, as many small farmers moved to non-farm employment, either in Bangkok or rural areas. Although some of the land formerly farmed by these households is no longer used for agricultural purposes (total agricultural area fell 6.6 percent), many households are now renting out their land to other farmers — the percentage of medium-size holdings (those between 6 and 22 hectares) that are rented in is now 41 percent, compared to just 23 percent in 1993 (the proportion of land rented in increased in all size classes).

Small farms in Suphan Buri increased their reliance on vegetables, livestock and fish during this period — the percentage of small farms engaged in these activities increased from 3 percent to 11 percent during those ten years. On the other hand, medium-size farms became more reliant on rice. Specialization in rice has been facilitated by the availability of water that has allowed farms to greatly intensify cropping — the percentage of rice farms growing two or more crops of rice in a year increased from 12 percent in 1987 to 53 percent in 1998 (Isvilanonda et al., 2000). Both types of farms are becoming more commercial (all rice production on medium-size farms is marketed outside the home), with each specializing in a different type of activity. This illustrates how the farming sector as a whole is diversifying, but certain classes of farms are specializing.

Farm households are also becoming older and more likely to be headed by females as young males migrate to urban areas for job opportunities. In Thailand, agricultural census data show that the proportion of farm households headed by women increased from 12 percent in 1978 to 27 percent in 2003. The same source shows that the percentage of farm household heads aged 55 and above increased from 25 percent in 1978 to 34 percent in 2003. This is not particularly surprising, as those who migrate from farms to cities in search of jobs are most likely to be young males (de Haan, 1999). These trends show that the clientele for water managers is already different from what it was in the past, and these trends are likely to continue in the future. Different clients are likely to mean that different strategies will be required to meet their needs.

4. Summary and conclusions

Compared to most other regions in the world, Southeast Asia is relatively rich in water resources. Nevertheless, the declining importance of agriculture and expanding populations mean that there will be increasing pressure to manage agricultural water resources more carefully. Improved incentives, institutions and technologies will be crucial to these efforts.

Because of crop diversification, farmers will increasingly demand irrigation that is more flexible (e.g. suitable for crops other than rice and cereals). For the many farmers who continue to plant rice, more reliable irrigation may be needed to prevent plant stress in the face of increased competition for water use from other sectors. The increased prevalence of pump irrigation is a response to these demands (Barker and Molle, 2004), and it is creating challenges for surface water managers to manage water conjunctively.

While demands for flexible and reliable irrigation increase, farmers will be less willing to devote their own scarce time to manage irrigation, because of the increased importance of non-agricultural income in their livelihoods, and will find it difficult to hire labourers to do it for them because of increasing labour scarcity and wages. This suggests that institutions that rely on substantial inputs of labour and time may experience difficulties, and new institutions that more directly confront the issue of incentives for individuals may have an important role to play in improved water management.

The clientele for water managers is already different from what it was in the past, and it will continue to change in the future. On average, the farmers of tomorrow will be older, more balanced in terms of gender, and will operate larger farms, although this latter characteristic will take longer to realize. Different clients mean that different strategies will be required to meet their needs.

References

Barker, R. & Molle, F. 2004. Evolution of irrigation in South and Southeast Asia. Comprehensive Assessment Research Report #5, International Water Management Institute, Colombo, Sri Lanka.

Dawe, D. 2005. Increasing water productivity in rice-based systems in Asia: past trends, current problems, and future prospects. Plant Production Science, Vol. 8, No. 3.

de Haan, A. 1999. Livelihoods and poverty: the role of migration — a critical review of the migration literature. Journal of Development Studies, 36(2): 1–47.

Fan, S. & Chan-Kang, C. 2003. Is small beautiful? Farm size, productivity, and poverty in Asian agriculture. Paper presented at the 25^th International Conference of Agricultural Economists, Durban, South Africa, August 2003.

Fan, S. Hazell, P. & Thorat, S. 1999. Linkages between government spending, growth and poverty in rural India. IFPRI Research Report 110, International Food Policy Research Institute, Washington, DC.

FAO. 2005. FAOStat on-line electronic database, accessed at http://faostat.fao.org.

FAO. 2004. The state of food insecurity in the world: monitoring progress towards the World Food Summit and Millennium Development Goals. Rome.

Hossain, M., Gascon, F. & Marciano, E. 2000a. Income distribution and poverty in rural Philippines: Insights from repeat village study. Economic and Political Weekly, Vol. XXXV, Nos. 52 and 53, December.

Hossain, M., Sen, B. & Rahman, H.Z. 2000b. Growth and distribution of rural income in Bangladesh: Analysis based on panel survey data. Economic and Political Weekly, Vol. XXXV, Nos. 52 and 53, December.

Huang, Q. 2004. Pump industry clusters in China, unpublished manuscript.

Isvilanonda, S., Ahmad, A. & Hossain, M. 2000. Recent changes in Thailand’s rural economy: evidence from six villages. Economic and Political Weekly, Vol. XXXV, Nos. 52 and 53, December.

Ministry of Agriculture, Forestry and Fisheries. 2004. The 78^th Statistical Yearbook of the Ministry of Agriculture, Forestry, and Fisheries, Japan.

National Statistical Office of Thailand. 1995. 1993 Agricultural Census. National Statistical Office, Ministry of Information and Communication Technology, Bangkok.

National Statistical Office of Thailand. 2005. 2003 Agricultural Census. National Statistical Office, Ministry of Information and Communication Technology, Bangkok.

Pingali, P. & Khwaja, Y. 2004. Globalization of Indian diets and the transformation of food supply systems. ESA Working Paper No. 04-05, Agricultural and Development Economics Division, FAO.

Rashid, S., Cummings Jr, R. & Gulati, A. 2005. Grain marketing parastatals in Asia: Why do they have to change now? Markets, Trade and Institutions Division Discussion Paper 80, International Food Policy Research Institute, Washington, DC.

Rosegrant, M.W. & Binswanger, H.P. 1994. Markets in tradable water rights: potential for efficiency gains in developing country water resource allocation. World Development, 22: 1613–1625.

Rosegrant, M.W., Cai, X. & Cline, S.A. 2002. World Water and Food to 2025. International Food Policy Research Institute, Washington, DC.

Sombilla, M.A. & Hossain, M. 1999. Rice and food security in Asia: a long-term outlook, Los Baños (Philippines): International Rice Research Institute. 38 pp.

Timmer, C.P., Falcon, W. & Pearson, S. 1983. Food Policy Analysis. Baltimore, Johns Hopkins University Press.

Vermillion, D.L. 1997. Impacts of irrigation management transfer: A review of the evidence. IIMI Research Report #11, International Irrigation Management Institute, Colombo, Sri Lanka.

Viet Nam General Statistical Office. 2000. Statistical Data of Viet Nam Agriculture, Forestry and Fishery 1975–2000. Statistical Publishing House, Hanoi.

Wang, J., Xu, Z., Huang, J. Rozelle, S., Hussain, I. & Biltonen, E. 2003. Water management reform, water use and income in the Yellow River Basin. In Proceedings of WCC-101, Agribusiness and food marketing in China, 17–18 April 2003 (available at http://www.china.wsu.edu).

World Bank. 2005. World Development Indicators on-line database (available at www.worldbank.org).

The World Conservation Union (IUCN) Asia Regional Water and Wetlands Programme, Bangkok, Thailand [email protected]

1.Introduction

IUCN should be a constructive change agent for more sustainable development, with a dual focus on ecosystems and livelihoods. In a world where governance is changing from a purely state-centric construct, the Union’s structure is well positioned to have more impact, now and in the future, giving extra space for both state and non-state actors to engage in political processes.

Many of us in the secretariat, in Asia and elsewhere, are particularly committed to water-related issues, which are increasingly prominent in Asia and world affairs. IUCN is a water actor in many different domains, such as: water infrastructure decision-making, flow regime negotiations, wetlands conservation, wetlands livelihoods, hydropower governance, irrigated agriculture, groundwater use, local adaptation to the impact of climate change and climate variability.

The IUCN Asia water team is committed to:

• water for nature — healthy ecosystems, especially rivers, lakes, wetlands, forests, fisheries, mountains;
• water for people — increasing livelihood opportunities and ensuring food security and sanitation as well as water’s contribution to spirituality; and
• water for development — sustainable national and regional development, especially employment, energy, transport, industry.

In Asia, the regional and national water team members have working relationships with state water-related players, local civil society organizations, and other international actors. There is much more to do, but the IUCN secretariat is committed to working with members and partners to improve the way in which we use water. This is the context within which we were pleased to accept an invitation to join the symposium organized by FAO in Ho Chi Minh, Viet Nam, 26 to 28 October 2005. As this brief written contribution was prepared after the event, we take the opportunity to congratulate FAO on their successful convening of a constructive, knowledge-sharing event.

2. Irrigation development

Southeast Asia

The growth of public funded and state-controlled large irrigation systems since the early twentieth century in Asia has been characterized by large water management projects that have harnessed rivers through the construction of dams, diversion structures, and canal systems. In the latter half of the twentieth century, the spread of irrigation technology has accelerated through state-sponsored large-scale irrigation and a greater emphasis on large dams for storage of river water. Irrigated areas worldwide increased from 40 million hectares in 1900 to ~100 million hectares by 1950 and then rapidly expanded further to ~280 million ha by 2000 (WCD, 2000). Most of this expansion in irrigated area took place in Asia. The irrigated area in Asia expanded from 65 million ha in 1950 to over 220 million ha in 2000. Asia now accounts for about 60 percent of the global irrigated area, about 60 percent of the global population, and about 90 percent of world rice production and 90 percent of global rice consumption (Maclean et al., 2002). In Southeast Asia the irrigated area has doubled between 1960 and 2000, from 9 million hectares in 1960 to about 18 million hectares in 2000. In Southeast Asia, rice is far and away the most important irrigated crop, accounting for on an average of 80 percent of irrigated cropping areas in the region, but as high as 98 percent in the case of Cambodia (see paper by Zhen in these proceedings).

A significant change in many countries has been the shift to have more multidisciplinary professional teams involved from the early stages of projects when developing irrigation systems. This is necessary if the mistakes of the past are to be avoided in the future.

Thailand

There are many examples in Thailand of irrigation systems that have been operating relatively successfully for a long period of time. However, an example of a poorly conceived and implemented project is the Khong-Chi-Mun (KCM) water diversion project, from which some lessons can be drawn.

The KCM project was originally planned to divert 6 850 million cubic metres of water annually from the Khong (Mekong) River by pumping and distributing it through a series of canals, aqueducts and storage reservoirs across vast areas of Northeast Thailand for irrigation. The projected cost of the project was over US$11 billion in 1992, and was to have taken 42 years to complete over three phases. Since the late 1980s, successive governments and senior politicians promoted the KCM project as a solution to the twin problems of water shortage and poverty in the Northeast, by utilizing the “wasted” waters of the Mekong to irrigate a target area of 4.98 million rai (7 970 km²) of rainfed agricultural land in the Mun-Chi Basins which feed the Khong. The project was first approved in 1989 and the Department of Energy Promotion and Development (DEDP), under the Ministry of Science, Technology and Environment was assigned responsibility.

Despite the completion of at least 12 dams along the Mun and Chi Rivers by the late 1990s, the KCM project has never met more than a fraction of its irrigation targets and some dams, like the Rasi Salai Dam in Sisaket Province were abandoned altogether, because of unpredicted environmental impacts and land conflicts which created social unrest and protests by local villagers and NGOs. Of particular concern was the rapid spread of salinization of water and farmland, rendering it unsuitable for agriculture and the loss of riparian forest to permanent flooding from reservoirs. Fisheries also declined precipitously around the same period, as the river was steadily fragmented from its mouth (Pak Mun) upstream. Although the KCM project has now been shelved since the dissolution of DEDP, the rationale and desire to complete a large pan-Isaan irrigation project is still alive, as evidenced by the recent re-rise (and fall) of interest in the northeast component of the so-called Thailand water grid mega-project.

The major food-bowl of present-day Thailand is the Chao Phraya Delta which has essentially been fully developed for irrigation, in a process spanning several hundred years since the Ayutthaya period (1350–1767), up to the present. The Upper Delta is supplied by a diversion dam at Chai Nat, whose water supply is partly dependent on the controlled release of waters from large storage dams upstream, such as the Bhumipol and Sirikit Dams. The Delta is the most populous part of Thailand and includes the capital Bangkok, with its associated industries and seat of political power, which compete with agricultural and domestic uses for a limited water supply. Most of the Delta has in the past been given over to double-cropping or even triple-cropping rice, although some parts are devoted almost exclusively to intensive vegetable or fruit tree cultivation, e.g. Damnoen Saduak to the west of Bangkok. It is estimated that during the dry season (January to June), when demand reaches its peak, available water supply is on average just above half the potential demand. The delta is marked by extreme hydraulic connectivity, which creates a high level of interactions and competition between the various users and actors, as well as its upstream water sources, and between surface and groundwater.

A recent study which took a political ecological approach (Molle, 2005) argued that the evolution of the Chao Phraya Delta has not been a linear process where technical change has allowed ever-increasing control of nature and development. Rather, the analysis argued that water use and further development has been highly politicized. Access to water has been constantly challenged and redefined, as old and new actors and interest groups have competed to obtain benefits. So also has it been with KCM.

These examples serve to remind us that the irrigation development must take into account ecosystems, livelihoods and governance. The next section offers some thoughts on these topics.

3. Ecosystems, livelihoods and governance

3.1 Ecosystems

Irrigation systems are made up of several components, e.g. reservoir, supply canals and irrigated fields that are artificial wetland ecosystems, often modifying or replacing earlier natural or semi-natural wetland ecosystems in the process of agricultural intensification and land conversion. Because of a poor understanding of the multifunctional uses of wetland ecosystems, these are often neglected in water management. A turn towards greater ecosystem consideration would seek to “strike a balance between benefiting from the natural resources available from an ecosystem’s components and processes, while maintaining an ecosystem’s ability to provide these at a sustainable level” (Pirot et al., 2000).

For example, where floodplain ecosystems exist in Southeast Asia, they are often associated with highly productive fisheries, which are dependent on natural flows and flood-drought cycles for their continued productivity. The best example in the region is the lower Mekong Basin, which is estimated to support a “wild” fishery yield of more than million tonnes per annum, which has been valued at US$1.2 billion (Sverdrup-Jensen, 2002). Upstream dams and impoundments in the Mekong Basin tend — as elsewhere — to impact negatively on wild fisheries because of flow alteration, nutrient capture, changes in water quality, loss of critical habitat, and blockage of fish migration routes.

In the lower Mekong Basin, there is finally an increasing interest in considering the needs of the wild fishery when considering further intensification of agriculture, energy production, and water flow regime changes.

To do so requires informed negotiations about how humans will intervene in natural flow regimes. Some toolbooks which aim to help different actors prepare for such negotiations have been prepared by IUCN, and may be of interest to the irrigation community assembled in Ho Chi Minh: “VALUE — Counting ecosystems as water infrastructure” (Emerton and Bos, 2004), “FLOW — The essentials of environmental flows” (Dyson et al., 2003), and NEGOTIATE (forthcoming).⁶

The point is to bring the ecosystems more explicitly into consideration, or take an “ecosystem approach” (Shepherd, 2004). In so doing, the full range of benefits and costs associated with irrigation systems are more likely to be taken into account.

3.2 Livelihood issues

Properly addressing the livelihoods perspectives of all the stakeholders in water allocation decisions, including construction and management of large irrigation projects, is critical since most irrigation projects are designed and justified in the name of poverty alleviation and food security. Therefore, irrigation projects should not jeopardize the livelihoods of any segment of society.

In the past decade or so, more and more bilateral donor institutions and development organizations have recognized the need to include sustainable livelihoods approaches (SLA) as an integral part of their work programmes and policy for addressing poverty reduction. SLA emphasizes understanding the vulnerability context and the organizational and institutional environment within which poor people draw upon assets of different types — human social, natural, physical, financial — to try and meet their needs (Scoones, 1998; Deardon et al., 2002).

SLA approaches provide a framework for fully taking into account the range of policy issues relevant to the poor early in the project planning cycle.

3.3 Governance issues

By water governance we mean the ways in which society shares power and negotiates with respect to decisions about how water resources are to be developed and used, and the distribution of benefits and involuntary risks from doing so. This includes the full spectrum of influences from shaping agendas and deliberating options through the design of institutions and laws, through the way these are implemented in the practices of day-to-day management of water. Governance is therefore not the privy of the state or confined to a particular scale or arena but emerges from the interactions between state, business and not-for-profit actors and their institutions at multiple scales (Lebel and Dore, 2005).

Southeast Asia has a long tradition of run-of-river farmer-managed irrigation schemes, and an even longer tradition of rainfed agriculture. More recently, large-scale schemes have been developed, and many more are being promoted and planned by state agencies. On the whole, much of this planning and construction takes place with minimal public information and consultation.

Governance analysis examines the rationales and processes for decisions about large-scale water infrastructure developments, such as large-scale irrigation schemes. These may, for example, involve interbasin diversions and the construction of storage dams, or be more about shifting responsibilities for water allocation or operations. Elsewhere, we are focusing our analytical work on the “re-packaged” different elements of the Thailand water grid and the irrigation works planned in northwest Cambodia (around Battambang), and northeast Cambodia (around Stoeng Treng), recognizing that there may be local, national and crossborder impacts. Through our analysis and engagement we expect to be able to get behind the discourses and document the beneficiaries of this “new water”. In doing so we also hope that such large regional water projects receive wide public scrutiny and are part of public debates and negotiations within and among countries.

4. Reflections and recommendations

New systems are being constructed, and existing systems modernized, in the developing countries of Southeast Asia, and so there is an urgent need to learn from past lessons and apply current best practice to ensure better outcomes and sustainability of the systems in the future. The following recommendations, generated in working groups at the Ho Chi Minh symposium, were offered for the consideration of those with an influence on irrigation design, construction and operation.

Comprehensive options and feasibility assessment

Workshop recommendation: Before committing to new, large-scale irrigation developments, a comprehensive options assessment should be made of the land and water existing use values and development options in that place. If a new, large-scale irrigation development is proposed, it should be examined by a wide-ranging feasibility analysis which is ecologically, physically, economically, politically, socially and culturally “logical”. These different logics should all be used to guide analysis and debate when examining the feasibility of a project. This should take place before progressing into the formal, legal, often rigid and relatively narrow “impact assessment” process. CSIRO’s 5-Way methodology and the WCD’s guidelines, where relevant, are international references.

CSIRO is an Australian science organization. The CSIRO 5-Way feasibility assessment methodology asks new projects to prove themselves as acceptable in five different ways: ecologically logical, physically logical (e.g. good engineering); economically logical (e.g. have the potential to provide financial benefit to society); politically logical; and socially logical. Of course, there will be different opinions as to what is logical or sensible in any place. Ideally these differences would be shared in the public domain and proponents of various views would be enabled to make their arguments.

WCD refers to the World Commission on Dams (WCD, 2000) which sought to undertake a global review of the development effectiveness of large dams, and assessments of alternatives. It wanted to create a framework for options assessment and decision making processes. It also wanted to identify internationally acceptable criteria and guidelines for planning, designing, construction, operation, monitoring and decommissioning of dams. The commissioners produced a “consensus” report, a negotiated opinion, which was launched in a blaze of publicity in 2000.

Source: CMU-USER summary extract from WCD report (Dore et al., 2004).

Figure 1. WCD framework for decision-making

Anticipate future changes

Workshop recommendation: If a new, large-scale irrigation development is proposed, the design must recognize and be flexible enough to take account of the inevitability of future demand changes. As economies improve and alter, land/water use and cropping systems will change. Therefore the function/service of the irrigation will change. From the initial stage of the development of an irrigation project, it is important to visualize the trajectory of how these changes might occur, e.g. from rice-focused production to more diversified enterprises.

This recommendation is a reminder that any comprehensive options assessment should also take into account possible futures.

Governance, water rights and responsibilities

Workshop recommendation: Large-scale irrigation projects, as with any others, should be planned, built and operated within a governance regime that embodies social justice ethics, is transparent, and participatory. Participation in irrigation governance should not be restricted to technical experts and bureaucrats, but should be open to representatives of affected communities and interest groups. The water rights and responsibilities of all stakeholders should be openly negotiated and established, with equity and sustainability being primary considerations. Management arrangements for a new project should include, from the beginning, credible representatives of different stakeholder groups.

This recommendation is a reminder that — in WCD parlance — there are global norms and some core values which should guide action.

Local capacity development

Workshop recommendation: If a new, large-scale irrigation development is proposed, it is essential to increase efforts to boost the capacity of local stakeholders playing many different roles. For example, local decision-makers need to be aware of the different options and feasibilities. Public authorities need to be skilled in designing terms of reference and overseeing contracts. Local consulting firms and engineers are required to construct and then be locally available to support ongoing operation, maintenance and adjustment. User groups need to be aided to improve water use efficiency. Local civil society organizations and universities should be able to play roles in governance, e.g. monitoring compliance with negotiated protocols and problem-solving. Supporting the development of this capacity needs to be factored into any new project.

Some of the discussants at the Workshop were very critical of past examples where local people and their institutions had not been factored into large-scale irrigation development and operation in ways that could have ensured more efficient management.

Finance

Workshop recommendation: In addition to the overall economic assessment, it is critical that an adequate financial strategy is put in place. The finance for complete construction must be ensured. Beyond construction, there must be a plausible strategy to ensure the availability of funds to meet full operation and maintenance costs, drawn from all project beneficiaries.

It is easy to see why this recommendation emerged. An over-emphasis in the past on anticipated net present value (NPV) or internal rate of return (IRR) and an under-emphasis on cash availability has left many schemes unable to be properly constructed, operated or maintained.

Monitoring impact on ecosystem and livelihoods

Workshop recommendation: Irrigation projects do more than supply water. They become part of the ecosystem and may have major impacts, for example on groundwater hydrology. The year-round effect of a project on the hydrology and wider environment have to be assessed; so does the impact, whether positive or negative, on the livelihood of all affected peoples.

This final recommendation of the Ho Chi Minh subgroup deliberating about new, large-scale schemes reflects a recognition that reductionist approaches in the past have led to insufficient analysis of the impact of large-scale schemes on nature and human development.

References

Deardon, P., Roland, R., Allison, G., & Allen, C. 2002. Sustainable livelihood approaches — from the framework to the field. Supporting livelihoods — evolving institutions. University of Bradford, UK, 29 to 30 May.

Dore, J., Lebel, L. & Manuta, J. 2004. Gaining public acceptance, Report for the UNEP Dams and Development Project. Chiang Mai University’s Unit for Social and Environmental Research, Chiang Mai.

Dyson, M., Bergkamp, G., & Scanlon, J. eds. 2003. FLOW — The essentials of environmental flows. IUCN, Gland, Switzerland.

Emerton, L., & Bos, E. eds. 2004. VALUE — Counting ecosystems as water infrastructure. IUCN, Gland, Switzerland.

Lebel, L., & Dore, J. 2005. M-POWER guide: operations manual for the Mekong Programme on Water Environment and Resilience. Unit for Social and Environmental Research, Chiang Mai University (available at http://www.mpowernet.org/).

Maclean, J., Dawe, D., Hardy, B., & Hattel, G. eds. 2002. Rice almanac: source book for the most important crop on Earth. Oxford, CABI Publishing.

Molle, F. 2005. Elements for a political ecology of river basin development: the case of the Chao Phraya River Basin, Thailand. In 4^th Conference of the International Water History Association. Paris, December.

Pirot, Y-J., Meynell, P. & Elder, D. 2000. Ecosystem management: lessons from around the world. A guide for development and conservation practitioners. The World Conservation Union (IUCN), Gland, Switzerland.

Scoones, I. 1998. Sustainable rural livelihoods: a framework for analysis. IDS Working Paper 72. International Development Studies, Sussex UK.

Shepherd, G. 2004. The ecosystem approach: five steps to implementation. IUCN — The World Conservation Union, Gland.

Sverdrup-Jensen, S. 2002. Fisheries in the Lower Mekong basin: status and perspectives. MRC Technical Paper #6. Mekong River Commission, Phnom Penh, Cambodia.

WCD. 2000. Dams and development: A new framework for decision-making. World Commission on Dams, Cape Town (available at http://www.wcd.org/).

⁶ VALUE, FLOW, NEGOTIATE (and others such as CHANGE, PAY, RULE and SHARE) are toolbooks in the IUCN Water and Nature Initiative series, and are downloadable from www.iucn.org/water