WATER MANAGEMENT IN RICE IN ASIA: SOME ISSUES FOR THE FUTURE - Thierry Facon^*

^* Water Management Officer, FAO Regional Office for Asia and the Pacific, Maliwan Mansion, 39 Phra Atit Road, Bangkok 10200, Thailand.

In most of Asia, rice is not only the staple food, but also constitutes the major economic activity and a key source of employment and income for the rural population. Water is the single most important component for sustainable rice production, especially in the traditional rice growing areas of the Region. Reduced investments in irrigation infrastructure, increased competition for water and large water withdrawals from underground water lower the sustainability of rice production. However, despite the constraints of water scarcity, rice production must rise dramatically over the next generation to meet the food needs of Asia’s poor. Producing more rice with less water is therefore a formidable challenge for the food, economic, social and water security of the Region.

This paper reviews the water resources and uses of the Region, the status of irrigation development with a particular focus in irrigated rice production, trends in the irrigation sub-sector but also in the water sector as a whole and in socio-economic development as they affect the sub-sector. The paper then examines water management practices for irrigated rice production and options and pre-set approaches to improve the water efficiency and productivity of rice production at the farm, and system level. These options must be considered in a basin-wide perspective and their adoption will require policy, economic and institutional reforms, as well as proper incentives, empowerment and irrigation services for farmers to adopt. Finally, the paper briefly describes the efforts and interventions needed to meet the challenge for producing more rice with less water.

2. IRRIGATION AND RICE PRODUCTION IN ASIA: AN OVERVIEW

2.1 Water Resources and Use^**

^** Data and tables in this section are drawn from the FAO Water Report 18, Irrigation in Asia in figures, 1999, published under FAO’s AQUASTAT Programme.

Overall, the Region is relatively well endowed with water resources. However, the amount of water resources per inhabitant is only slightly above half the world’s average. In terms of water resources per person, the groups of the Indian subcontinent, Eastern Asia and the Far East show the lowest figures while Southeast Asia has much more water resources per person than the world average. The figure of 2,000 m3/inhabitant/year is usually used as an indicator of water scarcity: India and China are reaching this limit, while the Republic of Korea is already below it, at 1,538 m3/inhabitant/year.

Table 1. Renewable Water Resources in Asia

Country	Population (1996)	Precipitation (mm)	Annual Renewable Water Resources					Dependency Ratio %
			Internal		External	Total
			million m3	m3 per inhab. 1996	million m3	million m3	m3 per inhab. 1996
	(1)	(2)	(3)	(4)=(3)*106/(1)	(5)	(6)=(3)+(5)	(7)=(6)*106/(1)	(8)
Bangladesh	120,073,000	2,320	105,000	874	1,105,644	1,210,644	10,083	91.3
Bhutan	1,812,000	4,000	95,000	52,428	0	95,000	52,428	0.0
Brunei	300,000	2,654	8,500	28,333	0	8,500	28,333	0.0
Cambodia	10,273,000	1,463	120,570	11,737	355,540	476,110	46,346	74.7
China	1,238,274,000	648	2,812,400	2,271	17,169	2,829,569	2,285	0.6
India	944,580,000	1,170	1,260,540	1,334	647,220	1,907,760	2,020	33.9
Indonesia	200,453,000	2,700	2,838,000	14,158	0	2,838,000	14,158	0.0
Japan	125,351,000	1,728	430,000	3.430	0	430,000	3,430	0.0
Korea, DPR	22,466,000	1,054	67,000	2,982	10,135	77,135	3,433	13.1
Korea, Rep.	45,314,920	1,274	64,000	1,431	4,850	69,700	1,538	7.0
Lao PDR	5,035,000	1,600	190,420	37,782	143,130	331,550	66,181	42.9
Malaysia	20,581,000	3,000	580,000	28,183	0	580,000	28,183	0.0
Maldives	263,000	1,883	30	114	0	30	114	0.0
Mongolia	2,515,000	251	34,800	13,837	0	34,800	13,837	0.0
Myanmar	45,922,000	2,341	880,600	19,176	165,001	1,045,601	22,769	15.8
Nepal	22,021,000	1,500	198,200	9,000	12,000	210,200	9,545	5.7
Papua New Guinea	4,400,000	3,500	801,000	182,045	0	801,000	182,045	0.0
Philippines	69,283,000	2,373	479,000	6,914	0	479,000	6,914	0.0
Sri Lanka	18,100,000	2,000	50,000	2,762	0	50,000	2,762	0.0
Thailand	68,703,000	1,485	210,000	3,577	199,944	409,944	6,983	48.8
Vietnam	75,181,000	1,960	366,500	4,875	524,710	891,210	11,854	58.9
Total	3,030,900,920	1,194	11,592,410	3,825

Water withdrawal: Table 2 shows the distribution of water withdrawal between the three major sectors of water use: agriculture (irrigation and livestock), communities (domestic water supply) and industry. Water requirements for energy (hydropower), navigation, fisheries, mining, environment and recreation, although they may represent a significant part of the water resources, have a negligible net consumption rate.

Table 2. Water Withdrawal in Asia

Country	Annual Water Withdrawal									% of intern. renewable water res.	% of total renewable water res.
	Year	Agricultural		Domestic		Industrial		Total
		million m3	% of total	million m3	% of total	million m3	% of total	million m3	m3 per inhab. (1996)
		(1)	(2)=(1)*100/(7)	(3)	(4)	(5)	(6)	(7)=(1)+(3)+(5)	(8)=(7)*100/(1) of T.1	(9)=(7)*100/(3) of T. 1	(10)=(7)*100/(6) of T.1
Bangladesh	1990	12,600.00	86	1,704.32	12	332.16	2	14,636.48	122	13.94	1.21
Bhutan	1987	10.80	54	7.20	36	2.00	10	20.00	11	0.02	0.02
Brunei	1994	-	-	-	-	-	-	91.60	305	1.08	1.08
Cambodia	1987	489.00	94	26.00	5	5.00	1	520.00	51	0.43	0.11
China	1993	407,774.00	77	25,165.00	5	92,550.00	18	525,459.00	424	18.68	18.57
India	1990	460,000.00	82	25,000.00	5	15,000.00	3	500,000.00	529	39.67	26.21
Indonesia	1990	69,241.00	93	4,729.00	6	376.00	1	74,346.00	371	2.62	2.62
Japan	1992	58,600.00	64	17,000.00	19	15,800.00	17	91,400.00	729	21.26	21.26
Korea, DPR	1987	10,336.00	73	1,557.60	11	2,265.60	16	14,160.00	630	21.13	18.36
Korea, Rep.	1994	14,877.00	63	6,209.00	26	2,582.00	11	23,668.00	522	36.50	33.96
Lao PDR	1987	812.00	82	79.00	8	99.00	10	990.00	196	0.52	0.30
Malaysia	1995	9,750.00	77	1,342.00	10	1,641.00	13	12,733.00	619	2.20	2.20
Maldives	1987	0.00	0	3.32	98	0.05	2	3.37	13	11.23	11.23
Mongolia	1993	227.04	53	85.36	20	115.72	27	428.12	170	1.23	1.23
Myanmar	1987	3,564.00	90	277.20	7	118.80	3	3,960.00	86	0.45	0.38
Nepal	1994	28,702.00	99	246.00	1	5.00	0	28,953.00	1,315	14.61	13.77
Papua New Guinea	1987	49.00	49	29.00	29	22.00	22	100.00	23	0.01	0.01
Philippines	1995	48,857.00	88	4,269.00	8	2,296.00	4	55,422.00	780	11.57	11.57
Sri Lanka	1990	9,380.00	96	195.00	2	195.00	2	9,770.00	540	19.54	19.54
Thailand	1990	30,200.00	91	1,496.00	5	1,436.00	4	33,132.00	564	15.78	8.08
Vietnam	1990	47,000.00	86	2,000.00	4	5,330.00	10	54,330.00	723	14.82	6.10
Total		1,212,469.64	84	91,420.00	6	140,171.33	10	1,444,152.97	476	12.46

In Asia, almost 84 percent of the water withdrawal is used for agricultural purposes, compared to 71 percent for the world. The Indian subcontinent and Eastern Asia have the highest level of water withdrawal for agriculture with 92 and 77 percent, respectively. The two regions together represent about 82 percent of the total irrigated area in Asia. With a major regional emphasis on flooded rice irrigation, it is particularly difficult to assess agricultural water use. The gross average for the Region is 8,900 m3/ha/year. Figures for China and India, which represent 72 percent of the Region’s agricultural water withdrawal, are relatively similar: 7,500 and 9,200 m3/ha of irrigated land, respectively. However, other countries show much higher values, as is the case for the Philippines, Malaysia, Japan, Republic of Korea, Nepal and Sri Lanka, where agricultural water withdrawal is between 15,000 and 31,500 m3/ha/year. More research is needed to obtain homogenous information on agricultural water use among countries.

Water withdrawal expressed as a percentage of Total Renewable Water Resources, which takes into account the incoming or border flows and the existing agreements, is a good indicator of the pressure on water resources. Roughly, it can be considered that pressure on water resources is high when this value is above 25 percent, as is the case for India and the Republic of Korea with 34 and 26 percent respectively. China, Japan, DPR Korea and Sri Lanka also have high values with 18.57, 21.26, 18.36 and 19.54 percent, respectively.

2.2 Irrigation

Irrigation potential: The irrigation potential for the Region was estimated at 235 million ha. India and China account for about 76 percent of this total. However, figures presented here should be used cautiously. In India, for example, the irrigation potential, which is 113.5 million ha, corresponds to the gross area which could theoretically be irrigated in a year on the basis of the assumed design cropping pattern and a rainfall probability of 75 percent, and represents 2.27 times the area under irrigation in 1993. This figure is a theoretical maximum. Indeed, it is considered that development of irrigation in India is about to reach its limits and that no major extension of irrigated lands is to be expected after the beginning of the twenty-first century. In China, the figure for irrigation potential is 64 million ha and corresponds to the total area which could be brought under irrigation in the first half of the next century. As much of the additional land proposed for irrigation is located in the arid and semi-arid zones, reaching such a level would require a viable long-term strategy as to how to provide the amount of water necessary to irrigate these lands.

Irrigation development: Asia represents the bulk of irrigation in the world. High population density combined with the tradition of irrigated rice cultivation in all the tropical part of the Region are the main factors explaining the importance of irrigation in Asia. While irrigation development dates back several centuries, the twentieth century, and particularly its second half, has seen a rapid increase in what could be called modern irrigation development and a majority of the countries have achieved self-sufficiency in cereal crops, mostly rice.

The assessment of land under irrigation in the countries of the Region is made particularly difficult by the different approaches used in the countries to compute irrigation. For some countries (Bangladesh, Bhutan) paddy fields, cultivated mainly during the wet season, are not considered as irrigated land. For the other countries where paddy rice cultivation is practiced, all paddy fields are considered irrigated land. In most cases, schemes are designed primarily to secure rice cultivation in the main cropping season, although the need for intensification has progressively led some countries to design new irrigation schemes for year-round irrigation, e.g. Thailand, while Vietnam has three rice crops a year. In total, 37 percent of the land under cultivation in the Region is irrigated.

While most wet season rice irrigation is fully gravity irrigation (cascades from plot to plot), dry season cropping may require pumping in places. In the tropical zone, wet season irrigation is almost only paddy rice. It is usually considered as supplementary irrigation to an already abundant precipitation. During the dry season, a much larger diversity of crops are grown on irrigated fields. In Cambodia, Indonesia, Malaysia and Mongolia, a kind of flood control irrigation is practiced with flood water being used to inundate paddy fields which are then cultivated with rice. In total, such practice concerns an area of about 1.2 percent of the total irrigated land in the Region. Surface irrigation is by far the most widespread irrigation technique in the Region and all rice is irrigated by surface methods. Surface water is the major source of irrigation water in the Region, except for Bangladesh, China, India and Pakistan where groundwater is widely used. The percentage of power irrigated area is more important in Bangladesh, China and India, with 83, 54 and 53 percent, respectively.

Irrigated crops: Rice represents about 45 percent of all irrigated crop areas in the Region and 59 percent of the rice is irrigated. However, its regional distribution shows major trends: in the countries of the Far East, Southeast Asia and the Islands, rice represents systematically more than 90 percent of irrigated crops, as is also the case for Bhutan, Nepal and Sri Lanka. By contrast, India, China and DPR Korea have a much more balanced distribution of irrigated crops with rice representing only about one-third to one half of the total irrigated crop area. This reflects the cold or arid context of large parts of these countries. In India, the percentage of land under irrigated wheat is slightly higher than that under irrigated rice (31 percent as against 30 percent). In China, it can be estimated that it is shared evenly between rice, wheat and other crops; rice being the single most important irrigated crop. However, in India only 47 percent of the total harvested area for paddy rice is irrigated, while more that 92 percent of the harvested paddy rice in China is irrigated.

Cropping intensity varies from 72 percent in Bhutan, to 132 percent in India and Malaysia with an average of 127 percent. Care should be taken, however, when comparing figures for different countries. In Bangladesh (84 percent), irrigation is considered only for dry season cropping. The average irrigated cropping intensity for ten countries where data are available is 127 percent.

There are approximately 28 million hectares under intense irrigation, producing two to three crops per year. Average yields are 4-6 t/ha per crop, and on a yearly basis 10-15 t/ha are common. Maintaining and improving the high annual output from these areas is essential for food security. However, there are signs of declining productivity in the intensively cropped, irrigated system, both on long-term research plots and farmers’ fields. Reasons for this phenomenon are not understood yet, but are thought to be linked to the prolonged submergence of soils, puddling and their effects on soil chemical and biological processes, including anaerobic decomposition of organic matter, and near-continuous soil reduction. Finding solutions to arrest declining productivity will therefore most certainly require changes in flooding practices.

2.2.1 Drainage, flood control and environmental issues

In most of Asia, drainage is closely linked to irrigation. In traditional terraced paddy cultivation, water flows from one plot to another and no distinction can be made between irrigation and drainage. In several humid countries of the Region, large segments of lowland or wetland are used for paddy cultivation. In such cases, the main purpose of water control is to ensure appropriate control of water level and drainage. Bangladesh and Cambodia use the terms controlled flooding or inundation, which are typical of paddy cultivation in the major deltas (Brahmaputra, Mekong). Lao PDR prefers to use lowland flooded rice. In these areas, drainage and flood control are also very much related. In Bangladesh, on average, 22 percent of the country is flooded every year and 50 percent of water development expenditures are spent on flood control and drainage. In Myanmar, in the Ayeyarwady Delta, drainage and flood control structures are also linked. Drainage covers 1 million ha in north and central Vietnam, mostly in the Red River Delta. Flood protected areas in China represent 32.69 million ha. The extreme case of agriculture under flood conditions is floating rice in Cambodia.

Drainage infrastructure associated with irrigation in arid and semi-arid areas concern mostly northern China, India and Mongolia. In China as a whole, it was estimated in 1996 that 24.58 million ha were subject to waterlogging, of which 20.28 million ha were equipped with drainage. In India, drainage works have been undertaken on about 5.8 million ha, but investment in drainage works associated with irrigation schemes has been widely neglected and drainage systems are usually in very poorly maintained condition.

Although total water withdrawal remains limited compared to water resources in Southeast Asia (about 5 percent), the large amounts of water diverted, mostly for agriculture, in those countries, have an environmental impact which may assume important proportions locally. Intrusion of saltwater in deltas is a concern in Myanmar, Vietnam and parts of India. Excessive groundwater exploitation around Bangkok, Thailand creates land subsidence and exacerbates already existing flood problems.

2.2.2 Trends in irrigation and drainage

Overall, growth in irrigated areas in Asian countries has declined from 2.1 percent per year in period 1961-1980 to 1.3 percent per year in 1980-1995. This decline is most acute in industrialized East Asia, followed by China. Most of the growth has come form tube-well development, especially in India, Pakistan and Bangladesh. However this groundwater development is not sustainable in many regions where groundwater draw-down has reached alarming levels, with very severe ecological impacts. A large proportion of new areas are not planted with rice but with other crops. Declining prices of rice, higher marginal development costs, environmental concerns, and poor performance of existing schemes are among the main factors for the decline in irrigation growth and investment both by governments and farmers in the Region.

However, the proportion of rice area that is irrigated is increasing, rising from just 35 percent of the total rice area to 44 percent over the last twenty years. Rice irrigated areas have expanded by 600 000 hectares per year while upland and deep water rice ecosystems decreased by 25 percent.

While irrigation has been instrumental in achieving self-sufficiency in staple crop production in recent decades in most countries of the Region, some countries such as Indonesia and the Philippines still indicate self-sufficiency as a major target of their irrigation development programmes; this is mainly to keep pace with rising populations. In Malaysia, however, the national policy is to decrease self-sufficiency in rice from 80 to 65 percent in 2010, due to the high cost of rice production. In Japan, rice irrigation has been on a downward trend for the last 20 years due to overproduction in the 1970’s.

Increased competition for water between sectors already affects agriculture in China, India, Malaysia, Thailand and the Republic of Korea and the trend is towards an intensification of the problem due mainly to the rapid growth of the domestic and industrial sectors in these countries. Major interbasin transfer programmes are reported in China and Thailand. Water scarcity and the interdependency between water use sectors are pushing countries to develop integrated water resources management programmes. Water quality and the increased importance of water conservation and protection are also major growing concerns.

The failure to develop adequate operation and maintenance (O and M) mechanisms to ensure the sustainability of the irrigation schemes (mostly large, public schemes) has led to irrigation management transfer or increased participation of users in the management of the schemes. This is achieved through the development or improvement of water users associations (WUAs).

Most of the countries have undergone deep societal and socio-economic transformations, characterized by: fast economic growth (until recently at least), especially in the industrial and services sectors; liberal macro-economic policies, development of trade reforms and privatization in the public sector and institutions; development of the civil society; and growing awareness of environmental issues and problems. In general, it is estimated that these profound changes in the environment, dominated by the need to adapt to water scarcity chiefly by the adoption of demand management strategies, call for a deep transformation of the irrigation sub-sector by the adoption of the following measures. First and foremost to consume less water, to modify water demands and maximize efficiency in water use and to improve of it’s economic, technical, and environmental performance, together with diversification of produce and cropping patterns, changes in management systems and structures, and financial and fiscal sustainability.

On the other hand, improved levels of education and of technological environment, more dynamic markets and diversified financing systems, more efficient and decentralized administration, and new management models, constitute many favourable conditions for an improvement of the performance of the irrigation sub-sectors and modernization of irrigation schemes.

As the older public schemes reach the age of 30-40 years in most countries, the issue of rehabilitation, which is related to operation and maintenance and modernization, is becoming increasingly important. While for some countries (such as Lao PDR, Myanmar, Philippines, Vietnam and parts of India) the extension of irrigated land still represents an important part of irrigation programmes, in most countries rehabilitation programmes are taking on increasing importance. The increased land and water scarcity and low expected return of future expansion of irrigation in these countries are often factors explaining the growing importance of rehabilitation in irrigation programmes.

Modernization of irrigation schemes as a part of a broader transformation of the water and agricultural sectors, responds to a complex set of institutional, technical, operational and economic issues, and would consist of a complex set of institutional, technical, operational and agricultural changes, generally associated with changes in water pricing and cost recovery. There is a general agreement on the specific objectives of the improvement of the performance of irrigation systems, in terms of delivering water to farmers in a more efficient, flexible, reliable and equitable manner. However, progress in the Region is rather slow when compared with other regions, and particularly with countries like Mexico or Turkey. Concepts related to service-oriented irrigation are not yet widespread or understood.

3. SCENARIOS FOR 2025

In World Water Demand and Supply, 1990 to 2025: Scenarios and Issues (IWMI Research Report 19, Seckler, Amarasinghe, Molden, de Silva and Barker, 1998), IWMI projects growth in water demand with two scenarios for the irrigation sector. In the first scenario, the 1990 level of irrigation efficiency remains constant through 2025. In the second scenario, higher efficiencies are attained (70 percent except for rice growing countries where 60 percent is projected, or a doubling of present efficiency, whichever is lower). The assumptions are that the per capita amount of food production from irrigated agriculture will remain constant. No allowance is made for additional irrigated area or irrigation water to meet increased per capita food demand; increased per capita consumption is met by increased yields; and the proportion of food supplied by irrigated areas and rainfed areas remains constant. These assumptions may underestimate the severity of the problem as cereal yields are stagnating, remaining rainfed land in the Region is very limited or can be developed only at high environmental and economic costs, and irrigation land is being lost to urbanization, water logging and sanitation at a fast pace.

Table 3 presents 1990 per capita withdrawals of water for the domestic, industrial and irrigation sectors and projected withdrawals by these sectors in 2025 for Asian countries. For countries currently below 20 m³ per capita for satisfaction of basic domestic water needs, 20 m³ per capita are projected in 2025. For countries currently above that level, estimates of withdrawal for domestic and industrial sectors are based on projected per capita GNP for both the domestic and industrial sectors. Environmental needs (minimum flows, water demand of ecosystems, etc.) are not taken into account. The model does not take into account shifts in production patterns, the role of trade in meeting national food balances, the trans-boundary nature of water resources and changes in food consumption patterns which usually follow socio-economic development. In this respect about 2,200 m³ of water are required to feed one person for a year with a diet rich in meat. A diet low in meat requires about half as much water.

Table 3. Water Supply and Demand in Asia in 2025

For the Region as a whole, in the first scenario total water withdrawal from all sectors would increase by 62 percent as against 18 percent in the second scenario, or a total saving of 691KM3 per year. Still additional withdrawals of 315 KM³ per year over the present withdrawal of 1,555 KM³ per year would be required and, globally at the regional level, potential water savings derived from increases in irrigation efficiency could not compensate for the growth in food and other demands. Additional water resources would need to be developed and there could be no net transfer of water resources from irrigation to the other sectors.

At the country level, countries can be grouped according to the nature and degree of their projected water scarcity by 2025 under the second scenario:

Group 1: absolute water scarcity^*** (Pakistan, Afghanistan and Singapore, total population 333 million). Singapore is a very particular case and must be treated separately as there is no irrigation in this City State. They do not have sufficient water resources to meet reasonable per capita water needs and will certainly have to reduce the amount of water used in irrigated agriculture and transfer it to other sectors, importing more food instead. This will place an additional burden on their economies as they are already suffering large deficit accounts. Being so large and diverse, China and India must be treated separately at a sub-national level. North China as well as West and South India are very dry and around one third of the population of these two countries will live in regions of absolute water scarcity (381 and 280 million people, respectively from a total of 661 million people).

^*** If annual withdrawals are higher than 50% of annual available resources.

Countries in Group 3 (Nepal, Australia, Cambodia, Myanmar, Malaysia) also need to increase water development by between 25 and 100 percent. They represent a total population of close to 500 million people and their capacity to make the necessary investments is very diverse. Countries in Group 4 (Philippines, Vietnam, Bangladesh) would have only modest requirements for additional water resources development while countries in Group 5 (Republic of Korea, DPR Korea, Japan, Thailand and Sri Lanka) would have zero or negative needs for water development.

The model has the merit of comparing two scenarios: business as usual and substantial (perhaps over-optimistic under any circumstances) increases in irrigation sector water use efficiency, in the context of projected growth of the water demand of all sectors; and of taking into account not only the availability of water resources but also an estimation for their required further development. The overall picture for the Region may be described as the following: because of the projected increases in population and therefore food demand, irrigated food production will need to increase significantly. Demand from other sectors will also increase because of both projected economic development and increase in population.

Many countries in the Region (Group 4 and Group 5) would be able to meet total societal water demand for their socio-economic development at the cost of relatively limited further water development (and therefore limited environmental impact) and/or with their available water resources, provided that they embark on significant and far-reaching improvement programmes of water use efficiency in the irrigation sector. The potential benefits or problems averted would be greater for those countries with limited investment capacity such as Vietnam or Bangladesh, which otherwise would need to almost double their developed water supplies.

Countries in Group 3 would need to invest massively in both water development and improvement of their national irrigation systems in order to avoid water becoming an overriding constraint in socio-economic development and to meet food security objectives, but they have a varied capacity to do so.

However, approximately one billion persons or about a quarter of the Region’s population would live in countries or regions of absolute water scarcity with severe consequences for their rural (and urban) population and substantial impact on the agricultural sector, for which Governments would need to prepare the populations and assist them in finding employment and income generating activities in other economic sectors, and develop other sectors to be able to meet their food import bills in order to achieve food security. This situation may be mitigated to a certain extent by inter-basin transfers within China or India.

As many of the regions concerned are major production areas for vital cereal food production, it is foreseen by many experts that the need for these regions to import cereals could have severe consequences for the poor segments of the population in other countries, by raising their prices on the international markets. A major factor of poverty eradication in the past has been the reduction of food commodity prices thanks to the (irrigated) “green revolution”. In theory, a shift in global production patterns for crops with a high virtual water content from water-scarce regions to well water-endowed regions and countries could ensure the satisfaction of demand, but whether this will happen is far from certain.

What seems to be certain is that nearly all countries in the Region will need to invest considerable efforts and resources in a mixture of improved demand management of the water sector and interventions on the supply side. In addition to the required economic investments on the supply side, considerable investments entailed by an irrigation water management improvement programme or the institutional and social capacity of the countries in implementing the necessary reforms in the water sector as a whole or in the irrigation sub-sector would be required to achieve the very considerable improvements in water use efficiency postulated in the second scenario. These rising costs will be borne increasingly by the water users through a combination of pricing and cost recovery, pushing the prices of food commodities up, impacting in particular on the Region’s poor.

4. RICE WATER MANAGEMENT

Total water requirements and specific water use (m³/ha) for rice production under different ecologies can be roughly estimated on average (evapotranspiration 550-950 mm/crop, which is the water actually consumed by the plant) at:

- rainfed upland rice: 5500 m³/ha (evapotranspiration only) for 1.25 t/ha specific water use: 6.5 m³/kg

- rainfed lowland rice: 10,000 m³/ha (evapotranspiration + impounded rainwater) for 2.5 t/ha specific water use: 4.0 m³/kg

- irrigated upland rice: 10,000 m³/ha (evapotranspiration + supplementary irrigation) for 2.5 t/ha specific water use: 4.0 m³/kg

- irrigated lowland/deepwater rice: 16,500 m³/ha (evapotranspiration and full irrigation) for 4.5 t/ha specific water use: 3.7 m³/kg

Research, with some reason, has concentrated in the past on this ecology where the greatest potential gains could be achieved per ha and globally. Early research focused on ways to improve water productivity by developing improved varieties and improving agronomic management, then more recently on improving water use efficiency, and finally on improving water productivity (which considers yields or income per m³ of water consumed) at all levels.

Irrigation inflow requirements (the amount of water diverted into the system) can be subdivided into crop evapotranspiration (T), evaporation (E), seepage and percolation losses (S and P), and surface run off (SRO). Because quantities of water required for land preparation and soaking as well as for maintaining water level in the paddy fields and soil saturation are high, T may represent only a small portion of irrigation inflow requirements and therefore overall (system) irrigation efficiency or (farm) water use efficiency are typically quite low (in the range of 30 to 40 percent).

Typically, parts of seepage and percolation losses as well as surface runoff can be re-used, i.e., recycled within the system (RCL). Attention has focused more recently on the fate of seepage and percolation and runoff. If this water is reused within the system (recycling drainage water or with conjunctive use) for agriculture or other uses, or returned to the hydrological cycle for further use downstream for productive use, then this water cannot be considered as lost. In the upstream part of the river basin, reducing these “losses” might only result in dry or paper savings and in disturbing the established hydrological regime (reducing groundwater re-charge, affecting downstream users etc.). Further downstream, wherever this water flows into sinks (i.e., cannot be reused), flows into the sea or is too polluted or salinized to be reused, then, attempts at reducing these losses or recycling them within the system would result in real or wet water savings. Indeed, it may be argued that paddy fields perform similar hydrological functions to wetlands for groundwater re-charge, flood control and trapping silt, which could be valued. Some authors have even suggested that farmers might be subsidized to practice inefficient irrigation practices for groundwater re-charge.

In any case, it is now widely accepted that:

- A river basin perspective should be adopted with much more attention being paid to defining the boundaries of intervention (farm, system, basin). Substantial progress has been made in defining concepts and methodologies (water accounting, modeling, etc.) but available data, which are already woefully inadequate to assess the merit of interventions at the farm or system level, water abstraction and even cultivated and irrigated areas, are even more lacking for the adoption of integrated river basin approaches.

- More attention must be paid to water quality issues and particularly the release of pollutants (fertilizers and other agro-chemicals) and salt concentration.

They therefore contribute to increasing not only water use or irrigation efficiency but also to improving or sustaining water productivity. Indeed, water management methods which improve water use efficiency have been developed with a view to maintaining crop yields and actually, when implemented properly, lead to yield increases (in the range of 15-20 percent in China for intermittent flooding and other methods). It follows that, although it is correct and necessary to use rigorous concepts for efficiency and performance at system and basin levels, and to determine under various conditions the optimum combination of improved technologies and water management practices that can meet water demand with least water consumed and managing return flows to ensure system and basin level efficiency, in practice it is difficult to find water management techniques proposed for adoption at the farm level which do not simultaneously raise irrigation efficiency and water productivity.

The range of possible strategies and their effect on various components of irrigation inflow requirements can be summarized in the following Table 4.

Table 4. Practices and Strategies to Improve Rice Water Productivity

Practices	T	E	S and P	SRO	RCL
Developing improved varieties	x
Improving agronomic management	x
Changing schedules to reduce evaporation		x
Reducing water for land preparation		x	x	X
Changing rice planting practices		x	x	X
Reducing crop growth water		x	x	X
Making more effective use of rainfall			x	X
Water distribution strategies		x	x	X
Water recycling and conjunctive use					x

These various practices and strategies are presented and discussed in detail in SWIM Paper 5 (Guerra, Bhuiyan, Tuong, Barker, 1998) as well as in Barker, Dawe, Tuong, Bhuiyan and Guerra, 1998 and Klemm, 1998 from which the above table is drawn and will be summarized or commented on in the following section.

4.1 Increasing Water Productivity

Developing improved varieties: High yielding varieties (HYVs) have more than doubled rice water productivity (against T) over the last decades. Hybrid rice has successfully been introduced in transplanted systems. However, the direct seeding method which is gaining increased acceptance is limiting the adoption rate of the hybrid rice technology since the process requires the use of much more of the costly seeds of hybrid rice per hectare than does the alternative method of transplanting rice. Direct seeding of hybrid rice is not economical with current hybrid seed production technologies. The New Plant Type (NPT) has been developed by IRRI scientists with the goal of raising the yield potential of conventional rice varieties to about 12-15 t/ha. NPTs are targeted for direct seeding conditions in an irrigated ecology. Biotechnology could amend many abiotic and biotic constraints to sustainable rice production including drought stress and tolerance to adverse soils and cold temperature.

Improving agronomic management: Improving pest control and nutrient management and other technologies that enhance yields increase output per unit of water (T). It should be noted that IPM techniques were developed in the context of large schemes where water supply was considered a constraint. Efforts are currently under way to integrate on-farm water management, IPM, nutrient management with the improvement of crop management (Pilot projects of FAO’s Special Programme for food Security in Sri Lanka, Cambodia, Nepal, Bangladesh, and Pakistan).

Changing the crop planting date and making more effective use of rainfall: Both these strategies require changes in water resources or reservoirs and farm management strategies and good cooperation between system operators and farmers.

Reducing water use for land preparation: Practices include land leveling (which contributes to better utilization of variable rainfall early in the season, reducing weeds, reducing S and P, improving fertilization application efficiencies and improving the timeliness of land preparation etc.), reducing the land preparation period, puddling, management of cracked soils (losses can be reduced by measures that minimize crack development during the soil drying period through straw mulching and dry shallow surface tillage on crack formation during the fallow period, or by impeding the flow of water through these cracks), and dry tillage.

Changing rice planting practices: Wet seeding of rice uses about 20-25 percent less water than in traditional transplanted rice methods and drastically reduces labor for establishing the crop from 30-person days per ha for transplanting to 1-2 person days. Improved water management practices during crop establishment (the first 2 weeks from planting) are crucial to enhance the weed-suppressing advantages that can be achieved by early flooding of wet seeded rice. Expansion of employment opportunities and crop intensification have resulted in the replacement of transplanting by direct seeding. Dry seeded rice saves even more water especially during land preparation.

Reducing water use during crop growth: Intermittent flooding, maintaining the soil in sub-saturated condition, alternate drying and wetting (as developed in various provinces in China) can reduce water applied to the field by more than 40 percent compared with continuous submergence methods without affecting yields. Increases in yields by up to 20 percent are actually reported. Some variants of these water management methods allow for storage and maximum use of rainfall. Optimum use of rainfall during the rainy season can more generally save reservoir water and increase areas irrigated during the dry season.

Supplementary irrigation of rainfed lowland rice: Supplementary irrigation either for crop establishment or at critical growth stages, particularly flowering, can prevent yield depressions of up to 40 percent or even crop failure one year out of five for T. Aman (monsoon season) rice in Bangladesh.

Water distribution strategies: Reducing inequities in water distribution among tertiary canals or within tertiary canal blocks through various systems of rotation should contribute to achieving a more even distribution, reduce losses and provide water to large areas. However, rotation systems are difficult to establish in practice.

Water recycling and conjunctive use: Conjunctive use was developed on the Indian sub-continent principally to compensate for the lack of reliability, inequities in distribution, and rigidity of canal water distribution systems, which constitute many obstacles to the development of productive irrigated production systems. It allows flexibility in availability of irrigation water and secures against failures in water delivery. It enables farmers to reuse seepage and percolation losses from canals and fields. However, conjunctive use and recycling of drainage water were not developed primarily to enhance water productivity or overall system efficiency and are usually not considered in design manuals of most irrigation agencies. Their development in uncontrolled conditions have led in many areas to groundwater draw-down and salinity problems. However, they are standard features of modern design methods. Drainage recycling has been applied very successfully, for instance in the MUDA scheme in Malaysia.

Alternatives to flooding techniques: While Barker, Dawe, Tuong, Bhuiyan and Guerra address the domain of surface irrigation basin on-farm methods, Klemm also discusses pressurized irrigation methods. It should be noted that, indeed, in theory, and also in practice, rice (both upland and lowland) can be irrigated with overhead as well as surface methods, and, among these, not only flooding and related techniques but also furrow and other surface methods. These techniques have been developed mostly in Latin America for upland rice, in the United States of America and in the Mediterranean Region which faces severe water scarcity and, the region in China where large rice areas are under arid to semi-arid climates. With the development of new varieties and the improvement of agro-technical methods and practices, yield obtained under aerobic conditions reach the level of production as under flood irrigation. Good results are also achieved with sprinkler irrigation of lowland rice.

However, growing irrigated rice under aerobic conditions still faces severe constraints:

- Higher inputs for weed control

- Increased susceptibility to diseases

- Imbalance of soil nutrients

- More know-how required in on-farm water management

- Increase of investment and maintenance costs

- Deep ingrained traditions and social customs based on flood irrigation management.

Being confronted with this rather large number of problems, it is not surprising that farmers are reluctant to shift to more demanding water management techniques than flooding. However, considering the growing water scarcity and pressure on the irrigated sub-sector within the water sector and on agriculture by other sectors of society and overall economic development policies described in previous sections, there is no choice and farmers must be provided both with a conducive environment and a proper production tool, i.e. better performing irrigation services.

It is the responsibility of governments to develop such a conducive environment which can be briefly summarized as follows:

- Legal support at national level for land use and water resources management (establishment of laws);

- Legal support at district and community level for land use and water resources management (integration of customary laws and establishment of regulations and by-laws);

- Technical support in upgrading irrigation systems for efficient water distribution;

- Agricultural support in adapting agricultural practices to modified irrigation methods;

- Financial support to initiate community-managed credit-schemes; and

- Human resources development at district and community level (area-based water resources management and on-farm water management).

4.2 The System Level

Improvements in the operation and maintenance of rice irrigation schemes through rehabilitation of the deteriorated systems, improvement of irrigation infrastructure for surface irrigation, irrigation management transfer, modernization, combining to various degrees institutional, organizational and technical changes, have been attempted in the Region with variable degrees of success. Studies undertaken by the World Bank in recent years have evaluated the impact of irrigation projects.

Jones (1995) evaluated the design of rice project in the humid tropics and concluded, from the strong degree of resistance of farmers to new design standards and the level of anarchy and chaos observed on the schemes, that the more reticulated systems, capable of supporting on-demand water delivery, were not appropriate under these climates.

A more recent study (OED, 1996, Rice, 1997) assessed the agro-economic impacts of investments in gravity-fed irrigation schemes in the paddy lands of Southeast Asia, to determine whether and how the quality of operation and maintenance (O and M) services influences the sustainability of those impacts.

At four of the six sites, the areas supplied by the irrigation systems were significantly less than planned. Cropping intensities were also substantially lower than expected at three sites and falling at a fourth. Only one scheme had attained both its area and intensity targets. Paddy yields varied widely - between schemes and in comparison with expectations - but a weighted average for the wet and dry seasons at all the schemes was about 3.3 tons, or 85 percent of appraisal projections. However, farmers had not diversified out of paddy. Indeed, the concentration on paddy had increased. Output was between 32 and 73 percent of appraisal estimates for five schemes. The returns had also been driven down by the decline of the international price of rice.

Overall, agency and irrigator performance appeared to be substantially better than expected. Farmers cooperated to achieve at least basic O and M objectives regardless of the level of maturity of the formal organization. There were no substantial negative constraints on irrigated production attributable to poor performance in O and M. Those O and M operations that are essential to keep sufficient supplies of water flowing to the great majority of the fields were adequately carried out. The study also noted the dismantling of complex technological control systems installed in the 1980’s in favour of fixed structures that have no adjustments and structures that adjust automatically to changes in water levels; and the rejection by farmers of both rotations and gates. Rotations do occur, but they tend to break down under conditions of shortage, which is when they are needed most.

The main finding was that given that they offered poor economics and low incomes, these paddy irrigation schemes faced an uncertain future. Small holder irrigated paddy could no longer provide the basis for a growing, or even stable household economy, driving younger family members off the farms while older members who stayed behind concentrated on basic subsistence crops. Consequently, social capital would erode and O and M standards were likely to suffer. As economies expanded, irrigated paddy would not be able to compete with the incomes to be had from other employment opportunities. Improved O and M performance would not rescue them. The study made these recommendations:

- Sharpen the response to O and M failures by disaggregating O and M; identifying the poorly performing components; and dealing with disincentives specific to each, such as the tertiary gates that farmers below consider unfriendly.

- Simplify the infrastructure and operations technology by converting to fixed and automatic controls that need less human intervention and by supporting authorities who plan with the farmers to abandon equitable rotations by rationing water during emergencies.

- Promote the transfer of management to farmers and their WUGs judiciously by recognizing that organizing user groups pays off, but also accepting that immature WUGs cannot handle some management responsibilities.

- Improve household earnings by diversifying cropping systems and supporting research, extension, and marketing services keyed to specialty crops and integrated, high value farming.

Key findings were:

- While 15 of the 16 irrigation projects visited had some aspects of modernization, none of them could qualify as “modernized” irrigation projects.

- The partially modernized projects did not have the chaos and anarchy that has been widely documented in typical (non-modernized) irrigation projects.

- Several projects have been modernized to the point that the water conveyance operations and hardware were able to support functional water user associations, and in turn those water user associations were collecting sufficient water fees to pay for all or most of the O and M expenses.

- Water user associations of some form (parastatal or private sector) provide distinct advantages if they are properly empowered. The “social” WUAs that are developed for the purpose of providing maintenance and collecting water fees were consistently either weak or imaginary. The “business” WUAs that hired staff to distribute water and ran the water distribution similar to a business operation were often quite strong.

- Farmers and managers appeared to be satisfied with a level of water delivery service that simply eliminates anarchy and also provides “sufficient” water to farms. Such criteria are insufficient to support modern field irrigation hardware and management.

- Modernization efforts which emphasized computer programmes for predicting canal gate movements and water deliveries were generally ineffective (or worse).

- Modernization needs were split between hardware, management, and a combination of the two. All projects needed improvements in both hardware and management.

- Successful projects stress improved communications, focus on operational data rather than statistical data, and require a minimum of paperwork for operators.

- Simple hardware and operational changes could give immediate benefits - if people just knew about them. There is a huge lack of awareness of how to design irrigation systems that provide good service. Examples of simple potential improvements were:

· Re-orienting employees from statistical data collection to operations and focusing on results rather than process.

· Using weir flow on cross regulators rather than only orifice flow.

· Modification of turnout operations for improved flow control and measurement, including some physical modifications to the turnouts.

· Installation of re-circulation systems within the project to easily collect and reuse spill.

· Improved voice communication and mobility of operators.

· Remote monitoring of spill points, and subsequent adjustment of the head works for the pertinent canal. This can be done manually with radios or even over a reliable telephone network.

· More frequent adjustment of flow rates at the source of a project, based on meaningful data from throughout the project.

· Programmes for improved irrigation scheduling for field irrigation are doomed to failure unless the water delivery service is well controlled, reliable, and flexible -- which means most such programmes are doomed to failure.

- There is a very serious shortage of trainers and consultants who can provide focused and pragmatic training and design which properly incorporates both strategies and details of hardware and management modernization.

- Modernization is a slow and expensive process. Many modernization projects are under-funded with respect to the expectations.

- Overall, there is a lack of understanding of modernization strategies and how to implement them.

The study concluded the following on appropriate modernization strategies:

- Irrigation project proposals, at the onset, must clearly define:

- The desired service that will be provided at all levels within the system. This requirement needs more than a few sentences in a report. Performance-based design requires that substantial thought and resources be dedicated to this matter.

- The operational procedures which will be used to provide this desired level of service.

- The hardware and irrigation project game plan (strategy) that is needed to implement the proper operation.

- The vision for all modernization programmes must be on the water delivery service that is needed 30 years from now.

- Direct government contributions to O and M activities can realistically be reduced if the projects are first brought up to the point where reasonable water delivery service can be provided.

The Case of the Indian Sub-continent: Irrigation schemes in large parts of India and Pakistan have been built on the design logic of “protective irrigation”. The idea is to reach as many farmers as possible to protect them against crop failure and famine which would occur without irrigation in regions with erratic monsoon rainfall. Water available in rivers or reservoirs is spread thinly over a large area. The amount of water a farmer is entitled to receive is insufficient to cover the full water requirements on all his land for an average rainfall year. Management principles in India (Warabandi, Shejpali, crop sanctioning) all involve the problem of rationing scarce water in a supply based system where the objectives of individual farmers differ from those of the scheme management. Typical irrigation systems have very few control structures. Canals are run at full supply level or have to be closed in order to achieve equal distribution of water to ungated chak outlets and to avoid deposition of silt. Construction costs are low but maintenance costs are high in comparison to the low level of irrigation service. Protective irrigation systems have been able to mitigate the effects of severe droughts and are still the backbone of the agricultural economies in Pakistan and India. However, it is now becoming apparent that the design logic of these systems may no longer be adequate for modern productive irrigated agriculture in an increasingly global economy.

The most pressing problems include: low efficiency in water distribution and use, unreliable water delivery, widespread vandalism of structures, poor maintenance, waterlogging and salinity, and insufficient cost recovery. Farmers could cope with these inefficiencies and make full use of advances of the green revolution in cases were they had access to fresh groundwater. However in areas which are less fortunate, because of saline or insufficient groundwater, yields are stagnating or declining.

The level of service provided to farmers clearly would not allow them to adopt the water saving technologies and management practices for rice production described in previous sections.

Successful irrigation systems feature high yields, service oriented irrigation management and financial autonomy. They may be described as productive irrigation.

FAO’s position is that a resolute modernization of irrigation schemes, through a combined strategy of institutional, managerial and technological change with the objective to change from a supply to service oriented mode of operation, building on current economic trends, is the adequate strategic choice to the present economic and social environment.

Other options include maintaining the status quo, with an exacerbation of existing problems, or enforcement of the protective irrigation concept by irrigation authorities which would have to continue to be dependent on state subsidies: levels of irrigation service and agricultural productivity will remain low except where canal water is supplemented by private wells.

This view seems to be supported by the Government of India, which in the 1998 preface to the World Bank Irrigation Sector Report, called for “a total revolution in irrigated agriculture... with much more focus on the improvement of performance of existing irrigated facilities and provision of a client-focused irrigation service... a paradigm shift in emphasis... toward improving the performance of existing irrigated agriculture... a second revolution in irrigated agriculture is required now”.

5. CONCLUSIONS

The challenge to produce more rice with less water, economically and in ways that will be adopted by farmers in a context of reformed agricultural and water policies and integrated water resources management appears formidable yet is vital for the food security of the Region. This will require considerable investments in economic as well as human resources.

A range of options are available for increasing the productivity and efficiency of water in surface irrigated rice ecologies. More radical options departing from traditional systems are also available and may be required. Over the past decades, substantial gains have already been achieved and farmers have demonstrated that, provided that they are empowered, have the economic incentives and an adequate production tool and irrigation service, they could quickly adopt substantial changes in their water management practices. However, new institutional and technical approaches have had limited impacts in the field.

The most appropriate strategies to adopt will vary over time and space and will have to be designed carefully with the involvement of the farmers, but will need to be resolutely forward-looking and perhaps revolutionary. Identifying the policies, management practices and technologies needed at farm, system and basin level will require a multi-disciplinary approach, substantial investments in collection and analysis of new and relevant information and research, as well as constant evaluation of present approaches and practices.

REFERENCES

D. Seckler, 1996. The New Era of Water Resources Management: From “Dry” to “Wet” Water Savings, IWMI Research Paper 1.

T. Facon, 1997. Emerging issues in water management for rice, In: FAO Rice Information Vol.1.

T. Facon, 1997. Modernization of irrigation schemes, synthesis of country papers, In: FAO Water Report 12, Modernization of Irrigation schemes, past experiences and future options.

E.B. Rice, 1997. Paddy irrigation and water management in Southeast Asia, OED, the World Bank.

T. Facon, 1998. Irrigation modernization training programme, In: Proceedings of the Fifth International ITIS Network Meeting, IWMI, CEMAGREF, FAO, WALMI

R. Barker, D. Dawe, T.P. Tuong, S.I. Bhuyian and L.C. Guerra, 1998. The Outlook for water resources in the year 2020: challenges for research on water management in rice production, In: Proceedings of the 19^th Session of the International rice commission, FAO

L.C. Guerra, S.I. Bhuiyan, T.P. Tuong, R. Barker, 1998. Producing More Rice with Less Water from Irrigated Systems, SWIM Paper 5, IWMI.

W. Klemm, 1998. Saving water in rice cultivation, In: Proceedings of the 19^th Session of the International Rice Commission, FAO.

S. Mbabaali, 1998. Supply and demand for rice: a medium- and longer-term perspective, In: Proceedings of the 19^th Session of the International Rice Commission, FAO.

R.S. Padora, 1998. Genetic diversity, productivity, and sustainable rice production, In: Proceedings of the 19^th Session of the International Rice Commission, FAO.

E.L. Pulver and V.N. Nguyen, 1998. Sustainable rice production issues for the third millenium, In: Proceedings of the 19^th Session of the International Rice Commission, FAO.

D. Seckler, U. Amarasinghe, D. Molden, R. de Silva, R. Barker, 1998. World Water Demand and Supply, 1990 to 2025: Scenarios and Issues, IWMI Research Paper 19.

D. Seckler, D. Molden, and R. Barke, 1998. Water Scarcity in the Twenty-First Century, IWMI Water Brief 1.

FAO, 1999. Irrigation in Asia in figures, Water Report 18.

FAO, 1999. IPTRID, World Bank, Modern water control and management practices in irrigation, impact on performance, Water Report 19.