Water for agriculture is critical for food security. However, water for irrigation may be threatened by rapidly increasing nonagricultural uses in industry, households, and the environment. New investments in irrigation and water supply systems and improved water management can meet part of the demand. But in many arid or semiarid areas-and seasonally in wetter areas-water is no longer abundant. The high economic and environmental costs of developing new water resources limit supply expansion. Therefore, even new supplies may be insufficient. Whether water will be available for irrigation so that agricultural production can provide for national and global food security remains an urgent question for the world.
This brief examines the relationship between water and food production over the next 30 years using IFPRI's IMPACTWATER model. This global model simulates the relationships among water availability and demand, food supply and demand, international food prices, and trade at the regional and global levels (Rosegrant and Cai 2000 describes the methodology and data sources). The baseline scenario incorporates our best estimates of the policy, investment, technological, and behavioral parameters driving the food and water sectors. We then look at how faster growth in municipal and industrial (M&I) demand and slower investments in irrigation and water supply infrastructure would affect food production.
Water withdrawal and consumption in 1995 and 2025 are shown in the table below. Withdrawal refers to water removed from a source, some of which may be returned to it and reused. Consumptive use is the water withdrawn from a source and actually consumed or lost to seepage, contamination, or a "sink" where it cannot economically be reused. Under the baseline scenario, total global water withdrawals for agricultural, domestic, and industrial use are projected to increase 23 percent from 1995 to 2025. Projected withdrawals increase 28 percent in developing countries.
Global consumptive use of water will increase by 16 percent, the vast majority in developing countries, where consumptive use across all sectors will increase by 18 percent.
Non-irrigation water demand will increase by 62 percent worldwide, 96 percent in developing countries, and 22 percent in developed countries. Consumptive use of water for irrigation worldwide will grow only 3.9 percent, which is significantly lower than the 12 percent increase in the rate of water use to meet full demand for irrigated crops. Of critical importance, the slow growth in irrigation water supply, especially in developing countries, will be due to water supply constraints and high non-irrigation demand, which will increase water scarcity for irrigation.
Under the baseline scenario, world food prices will decline, but more slowly than in the past two decades. Wheat prices are projected to decline by 8 percent between the base year and 202125, maize prices by 7 percent, and rice prices by 17 percent. This compares to a decline of 3040 percent in these commodity prices between 1985 and 1995. The importance of developing countries in global food markets will increase substantially: 86 percent of the projected increase in global cereal consumption and nearly 90 percent of the increase in global meat demand between 1995 and 202125 will come from developing countries. Total cereal demand is projected to grow 43 percent by 760 million metric tons (mt).
Substituting food imports for irrigated agricultural production paid for by urban and commercial growth (so-called imports of "virtual water") is a possible strategy for reducing agricultural water use. However, even under the baseline scenario, developing countries' reliance on food imports will increase dramatically. With the slowing of crop yield growth, import demand in developing countries is projected to increase from 108 million mt in 1995 to 238 million mt in 202125. Irrigated and rainfed production will each account for about half of the increase in cereal production between 1995 and 202125. Irrigation plays a more dominant role in cereal production in developing countries, where nearly 60 percent of future cereal production will come from irrigated areas, accounting for four-fifths of the growth in global irrigated cereal production.
With area harvested projected to grow slowly, crop yield growth is essential to future food production. However, cereal yield growth in most countries will slow. The global yield growth rate for all cereals is expected to decline from 1.5 percent per year during 198295 to 1.0 percent per year during 19952025; and in developing countries, average crop yield growth will decline from 1.9 percent per year to 1.2 percent. Increasing water scarcity will be a primary cause of the slowdown in projected irrigated cereal yield growth in developing countries. The relative crop yield for cereals in irrigated areas in developing countries is projected to decline from 0.86 in 1995 to 0.74 in 2025. Relative crop yield is the ratio of the projected crop yield to the maximum economically attainable yields at specified crop and input prices under conditions of zero water stress. The fall in the relative crop yield index is a significant drag on future yield growth. For developing countries, the drop represents an annual cereal production loss of 139 million mt, which is higher than China's total rice production in 1995.
The baseline scenario, characterized by declining water supply reliability and relative crop yields, indicates the potential vulnerability of agricultural production to a worsening in water scarcity. What would happen if there were significant increases in non-irrigation water demand or declines in water investment and policy reform compared to the baseline? Under the high non-irrigation water (HNIRW) demand scenario that postulates faster growth in M&I demand, non-irrigation consumption will reach 681 km3 in 202125, 33 percent of total water consumption (versus 25 percent under baseline). Under the second alternative scenario, which assumes low investment in infrastructure (LINV), the net increase of global reservoir storage for irrigation and water supply rises by only 396 km3 between 1995 and 2025 compared to an increase of 690 km3 under the baseline. Global average basin irrigation efficiency increases only to 0.57 versus 0.61 under the baseline, corresponding to a water consumption savings of 23 km3 under the LINV scenario compared to 115 km3 under the baseline. The net increase in withdrawal capacity between 1995 and 2025 is only 301 km3 under the LINV scenario versus 844 km3 under the baseline.
The scenarios that project significant negative impacts on irrigation water consumption and irrigation investment result in declines in cereal production and large increases in cereal prices.
By 2025, irrigated cereal production will be 58 million mt less under HNIRW, and 103 million mt less under LINV than under the baseline. Rice prices are projected to be 30 percent higher than the baseline in 202125 under LINV and 15 percent higher under HNIRW. Maize prices are 22 percent higher under LINV and 27 percent higher under HNIRW, and wheat prices 22 percent higher under LINV and 19 percent higher under HNIRW. High non-irrigation water use and low investment also results in increased variability in international prices, which can pose adjustment problems for farmers. Moreover, the fall in irrigated area and production creates additional pressure on the relatively fragile rainfed land base. Due to rising prices, rainfed cereal area will increase by an annual average of 7 million hectares under the LINV scenario in 20212025.
At the local and regional levels, price increases of this magnitude would cause a significant decline in the real income of poor consumers. Malnutrition would increase because many of the poorest people in low-income developing countries spend more than half their income on food.
Investments and policy reforms in water and irrigation management will be significant determinants of future food production, demand, prices, and trade. Rapidly growing municipal and industrial water demand in developing countries will increase water scarcity for agriculture, and with a continued slowdown in water investments, could be a serious threat to future growth in food production. Food production, demand, trade, and prices will be affected. Three broad strategies can address the challenge of increasing water scarcity: (1) attack the problem from outside the water sector-for example, through crop breeding for more rapid yield growth; (2) increase the supply of water for food production through investments in irrigation, dams, or urban water supply; and (3) conserve water and improve the efficiency of water use through water management and policy reform. Appropriate water and food strategies must be countryspecific and will combine elements of each of these approaches.
A decline in water available for irrigation without compensating investments and improvements in water management and water use efficiency-in both irrigated and rainfed areas-will reduce production growth and sharply increase international cereal prices, causing negative impacts on low-income developing countries and consumers.
For further information see Rosegrant, M. W. and X. Cai. "Modeling Water Availability and Food Security: A Global Perspective: The IMPACT-WATER Model." Working paper. IFPRI: Washington, D.C., 2000.; Hofwegen, P. van and M. Svendsen. A Vision of Water for Food and Rural Development. World Water Vision Report, 2000.
Mark W. Rosegrant (m.rosegrant@cgiar.org) is a senior research fellow and Ximing Cai (x.cai@cgiar.org) is a postdoctoral fellow in IFPRI's Environment and Production Technology Division.