Can we raise grain yields fast enough?

Many countries have tripled or even quadrupled the amount of grain they produce per hectare of land. But now these gains are slowing, while global demand continues to soar. We need a new strategy for food security.

After a half-century of global surpluses of wheat, rice, corn, and other grains, it is easy to be complacent about the food prospect for the twenty-first century. We have come to take for granted the supply of grain that provides half of humanity's food energy when consumed directly and a good portion of the remainder when consumed indirectly in the form of livestock products.

But this complacency can be dangerous. Each year, as population continues to expand, the world's farmers must stretch their production capacity to feed an additional 80 million people. Beyond that, they must now satisfy the needs generated by record rises in affluence. As people make more money, they consume more beef, pork, poultry, milk, eggs, beer, and other grain-intensive products. A kilogram of pork, for example, may require four kilograms of grain to produce, so as people are able to afford more pork their demand for grain increases. And Third World incomes are now rising at record rates. In Asia, where more than half the world's people live, incomes are rising faster than they have on any continent at any time in history. This combination of more people and more consumption per person is putting heavy pressure on the land.

The world's farmers responded heroically to past increases in demand, nearly tripling the grain harvest from 630 million tons in 1950 to 1.8 billion tons in 1990. Most of this expansion came not from plowing a lot more land, but from more than doubling the amount of grain produced on existing farmland. Between 1950 and 1990, the yield per hectare grew at 2.1 percent per year. Augmented by whatever new land could be added to grain production, including that from expanded irrigation in arid regions, this boosted total grain production by an average of nearly 3 percent a year throughout that four-decade run - well ahead of population growth.

Although there were disastrous shortages from time to time during this period - in China, Ethiopia, and Somalia, for example - and although some 800 million people are still hungry and malnourished, overall supply has not been a major issue. In the United States, the government paid farmers not to plant part of their land. The steady growth in the harvest, and resultant decline in grain prices, created a psychology of surpluses - a psychology that has made it easy for policymakers both to put off the difficult task of stabilizing human population and to take their farmers' capacities to meet future challenges for granted.

The world's total demand for food is likely to nearly double its present level by 2030, and there is little new land available to plow. The key to food security in the years ahead, then, is whether farmers can continue to rapidly raise the productivity of their land, as they have done in the past. However, assessments of the potential for raising land productivity vary widely. In a recent World Bank report, researchers indicated that they expect grain yields to increase at 1.5 to 1.7 percent per year, or "at rates comparable to those in recent years. ..." With this rosy outlook, the Bank projects a surplus capacity in world agriculture as a whole, accompanied by declining food prices. This Worldwatch Institute analysis comes to a very different conclusion.

The World Bank economists base their projections on simple extrapolation, arguing that "historically, yields have grown along a linear path from 1960 to 1990, and they are projected to continue along the path of past growth." Although extrapolating past yield trends worked well enough in previous decades, it won't work in a world where the yields simply are not continuing to climb rapidly. In contrast to the robust increases of 2.1 percent per year between 1950 and 1990, the rise between 1990 and 1995 averaged only 1 percent a year. Although this period is too short to establish a clear trend, it may offer a strong indication of what the future holds.

Reliance on the World Bank projections by governments is leading to underinvestment in both agriculture and family planning. Funding for agricultural research is being cut by many governments, including that of the United States. At the international level, a striking example is the fate of the Philippines-based International Rice Research Institute (IRRI), which gave Asia the high-yielding rices. In 1996, several donor governments, facing cutbacks in their aid budgets, cut their general support funding of IRRI, the world's premier rice research institute, forcing a heavy, cutback in core staff. Similarly, in 1996 the U.S. Congress voted to cut fiscal year 1997 funding of international family planning assistance by 60 percent from 1995 levels, with little consideration of how this would affect the increasingly precarious balance between population and food supply. Fortunately, the new congress voted in early 1997 to restore part of the funding.

The Bank projections breed complacency, not urgency. They permit governments to treat prime cropland like a surplus commodity - one that can be paved over, built on, or otherwise frittered away with impunity. One result can be seen in California's Central Valley, where housing projects are marching up the valley unimpeded, consuming some of the world's finest farmland. In China, this process is taking place on an even larger scale, as the government paves over millions of hectares of cropland so the bicycle can be replaced with the automobile. And in Indonesia, fertile riceland is being converted to golf courses.

The central question now is whether farmers can restore the rapid rise in land productivity. Moreover, that question needs to be addressed in terms that point to realistic possibilities for farmers working under the natural constraints of their own environments - the availability of sunlight, water, and good soil. Rather than analyzing yields on experimental plots or those achieved by the best farmers, this analysis will assess the long-term yield potential under field conditions by individual countries.

The Century of Soaring Productivity

The first recorded case in which a country's farmers achieved a sharp increase in output per unit of land - a "yield takeoff" - began more than a century ago in Japan. In 1878, Japanese rice farmers got an average of 1.4 tons of grain per hectare. By 1984, the average yield had more than tripled, to 4.7 tons. Since then, it has plateaued - fluctuating between 4.3 and 4.6 tons in all but three years. [ILLUSTRATION FOR FIGURE 1 OMITTED] Despite the fact that Japan supports the price paid to its farmers for rice at four times the world level, thereby offering a powerful financial incentive to raise yields higher - and despite its ability to provide the best technology available - it has been unable to improve average yields for more than a decade.

In the United States, the first yield takeoff came more than a half century later, with wheat. During the nearly 80 years between the Civil War and World War II, U.S. wheat yields had fluctuated around 0.9 tons per hectare. As World War II got underway, and demand for U.S. grain rose as production was disrupted abroad, farmers began investing in higher-yielding seeds and in fertilizer. By 1983, yields had climbed to 2.65 tons per hectare, nearly tripling the traditional level. Since then, however, there has been no further rise. Although the wheat yield takeoff in the United States began decades after that of rice in Japan, farmers in the two countries appear to have "hit the wall" at about the same time. Two questions arise: Can scientists restore the historical growth in yields, or does this plateauing in two of the most agriculturally advanced countries signal a future leveling off in other countries, as farmers exhaust the principal means of increasing yields.

The Factors That Increase Yields

The 2.5-fold increase in world grain land productivity since 1950 has come from three sources: genetic advances, agronomic improvements, and some synergies between the two.

On the genetic front, most growth has come from redistributing the share of the plant's photosynthetic product (photosynthate) going to the various plant parts (leaves, stems, roots, and seeds) so that a much larger share goes to the seed - the part we use for food. On the agronomic front (where farmers' practices weigh heavily), advances that help plants realize their full genetic potential include the use of fertilizer, irrigation, the control of plant diseases and predatory insects, and the eradication of weeds.

Scientists estimate that the originally domesticated wheats devoted roughly 20 percent of their photosynthate to the development of seeds; they were stalk-heavy, harvest-light. Through plant breeding, it has been possible to raise the share of photosynthate going into seed - the "harvest index" - in today's high-yielding grain to some 50 to 55 percent. Given the plant's basic requirements of an adequate root system, a strong stem, and sufficient leaves for photosynthesis, scientists believe the physiological limit is around 60 percent.

One of the earliest gains in this area came in the late nineteenth century, when Japanese scientists incorporated a dwarf gene into both rice and wheat plants. Traditional varieties of these grasses were tall and thin, because their ancestors growing in the wild needed to compete with other plants for sunlight. But once farmers began controlling weeds among the domesticated plants, there was no longer a need for tall varieties. As plant breeders shortened both wheat and rice plants, reducing the length of their straw, they also lowered the share of photosynthate going into the straw and increased that going into seed. L.T. Evans, a prominent Australian soil scientist and plant physiologist who has long studied cereal yield gains and potentials, notes that in the high-yielding dwarf wheats, "the gain in grain yield approximately equals the loss in straw weight."

With corn, similarly, varieties grown in the tropics were reduced in height from an average of nearly three meters to less than two. But Don Duvick, for many years the director of research at the Pioneer Hybrid seed company, observes that with the hybrids used in the U.S. corn belt, the key to higher yields is the ability of varieties to "withstand the stress of higher plant densities while still making the same amount of grain per plant." One of the keys to growing more plants per hectare is to replace the horizontally inclined leaves of traditional strains that droop somewhat with more upright leaves, thereby reducing the amount of self-shading.

But while breeders can manipulate the distribution of photosynthate within the plant, the amount produced by a given leaf area remains unchanged from that of the plant's wild ancestors. Although plant breeders have greatly increased the share of the photosynthate going to the seed of the various grains, they have not been able to alter the basic process of photosynthesis itself.

On the agronomic front, the principal means of increasing land productivity have been to expand irrigation, use more fertilizer, and to more effectively control diseases, insects, and weeds. All of these tactics help plants reach more of their full genetic yield potential. Between 1950 and 1990, the amount of irrigated land in the world increased from 94 million to 240 million hectares, or 2.4 percent per year. Between 1990 and 1994, however, the official data show irrigated area increasing by only another 9 million hectares, or 0.9 percent per year. And because governments do not always report the land taken out of irrigation, some analysts doubt that there has been any net growth in irrigated area at all since 1990.

In the United States and China, the world's two largest grain producers, losses are all too visible - and in some instances perhaps irreversible. Texas, a major farming state that has historically relied heavily on irrigation, has lost 14 percent of its irrigated area since 1980 as a result of aquifer depletion. California, Kansas, and Oklahoma, too, are losing irrigation water. In China's Hebei Province, irrigation water is being diverted to cities to satisfy mushrooming urban and industrial demands for water. In the agricultural region around Beijing, farmers have not been allowed to draw water from the reservoirs since 1994, because all the region's water is now needed to satisfy the capital city's growing thirst.

David Seckler, head of the International Irrigation Management Institute in Sri Lanka, believes that world irrigated area actually may have started to shrink. If the future brings little or no growth in irrigated area, the world will have lost a major source of rising land productivity, since the expansion of irrigation also greatly expands the potential for using fertilizer.

Fertilizer helps to ensure that plant growth won't be inhibited by any lack of nutrients. With a ten-fold rise in fertilizer use, from 14 million tons in 1950 to some 140 million in 1990, this has been by far the most important agronomic source of higher land productivity since mid-century. But in the 1990s, use of fertilizer - like that of irrigation - has leveled off in many countries. U.S. farmers, after discovering that there are optimal levels beyond which further applications aren't cost-effective, are using less fertilizer in the mid-1990s than they were in the early 1980s. The leveling off in the United States has been followed by a similar trend in Western Europe and Japan. In the former Soviet Union, fertilizer use fell precipitously after subsidies were removed in 1988 and fertilizer prices climbed to world market levels.

Other agronomic contributions to higher cropland productivity include the more timely planting of crops made possible by mechanization and higher plant populations per hectare, the latter applying particularly to corn. More timely planting boosts yields because in the temperate zones there is typically a brief window of time for seeding, usually measured in days, when optimum yields can be obtained. If planting is delayed, then yields decline with each day of delay.

Advances in plant breeding and agronomy often reinforce each other. The dwarfing of wheat and rice plants not only reduced the amount of photosynthate that went for straw, for instance, but increased the benefit of adding more fertilizer. For example, the traditional tall, thin-strawed wheat varieties grown in India could effectively use only about 40 kilograms of nitrogen per hectare. Applications above that made the plants grow heavier heads of grain, but these would often "lodge," or fall over (especially in storms), leading to crop losses. With the dwarf varieties, however, farmers could boost nitrogen applications up to 120 kilograms per hectare or more, thus greatly increasing the yield, but with little fear of lodging. This synergy between genetics and agronomics helps to explain the doubling or tripling of yields achieved with the first generation of high-yielding wheats and rices that were at the heart of the Green Revolution.

With corn, the greater tolerance for crowding enabled growers to greatly increase the plant population - and hence the number of ears harvested - per hectare. At the same time, herbicides were being developed that would control weeds, eliminating the traditional need to plant corn rows far enough apart to permit mechanical cultivators to pass through the field during the earlier part of the growing season. As a result of these two advances, plant populations have climbed. In Iowa, for example, corn plant densities have nearly tripled since 1930.

For each of the three major grains - wheat, rice, and corn - the major worldwide gains in productivity took place between 1950 and 1990. Since 1990, gains have been much smaller, and the question now facing planners is just how much more can be expected.

Wheat Yield Potential: The Wide Variability of Natural Limits

Yields of wheat, which along with rice is one of humanity's two principal food staples, vary enormously from one country to another. Kazakstan, though it is the largest producer among the central Asian republics, produced only 0.6 tons per hectare in 1995, for example, whereas France, the largest producer in Europe, produced more than 10 times that - 6.8 tons per hectare. (See Table 1.) It may be tempting to cite such disparities as evidence of potential for improvement in the lower-yielding countries, but this would be highly misleading. In fact, the prospect for further raising yields may actually be better in France than in Kazakstan. Rainfall is marginal for agriculture in Kazakstan, and with wind erosion of soil now widespread, Kazak soil scientists report that cropland fertility is falling.

The fallacy of thinking that low yields are a measure of the potential for improvement is further illustrated by a comparison of India and Australia. In 1950, both countries were getting the same yields from their wheat: about 0.9 tons per hectare. By 1995, India had nearly tripled its yield, to 2.5 tons. Australia has increased to only 1.7 tons. But the difference is not a reflection of the superior capabilities of Indian farmers. In fact, the Australians had to use great ingenuity and effort to achieve the gains they got. The difference is that farmers in India, who irrigate much of their wheat, have good soil moisture, whereas those in Australia, who have to rely on sparse rainfall, do not. This also explains why Africa, a largely arid and semiarid continent, has not achieved the yield gains of Asia. With neither abundant rainfall or irrigation to provide sufficient soil moisture, farmers cannot effectively use much fertilizer - and therefore cannot fully exploit the genetic potential of their crops.

Table 1. Wheat Yield Per Hectare in Key Producing Countries, 1995(1)

Country                         Metric Tons

United Kingdom                      7.7
France                              6.8
Egypt                               5.6
Mexico                              4.1
China                               3.6

Poland                              3.4
Ukraine                             2.7
India                               2.5
United States                       2.5
Canada                              2.3

Argentina                           2.1
Pakistan                            2.0
Australia                           1.6
Russia                              1.4
Kazakstan                           0.6

1 Yield for 1995 is the average of 1994 through 1996. Source: USDA

Now that developing countries are collectively using more fertilizer than the industrial countries and, with few exceptions, using high-yielding varieties and other advanced technologies, ranking on the yield chart is determined largely by environmental factors, such as rainfall, daylength, and solar intensity. When a country's wheat-growing practices do fall short of their potential, the yield can be improved rapidly until environmental limits are reached - and then no amount of money, ingenuity, or fertilizer can take it much further. Both France and China, for example, were able to roughly quadruple their wheat yields over the past half-century. [ILLUSTRATION FOR FIGURE 2 OMITTED] But the United States, for all its world-leading access to technology, fertilizer, and agronomic knowledge, has not been able to match this and indeed has not been able to improve yields at all for the past 13 years. Since the United States is the leading wheat exporter, and the country on which the world's wheat-consuming countries rely for imports, this flatlining of yields may serve as a belated warning.

Raising wheat yields in China, particularly, is becoming more difficult as aquifers are depleted and the response to additional fertilizer diminishes, and as the country's explosively growing cities pull irrigation water away from agriculture. With the Yellow River now running dry for several weeks each spring, and for progressively longer periods each year, some wheat farmers downstream will face shrinking irrigation water supplies. For some weeks in the spring and early summer of 1996, the river ran dry before it reached coastal Shangdong Province, which until now has produced one-fifth of China's wheat.

In France, and in other major wheat-producing countries such as Germany and the United Kingdom, it took less than five years to go from 5 to 6 tons per hectare, but more than a decade to go from 6 to 7 tons per hectare. In almost every place where wheat is produced, in developing countries as well as the highly mechanized ones, the historic rise in yields is slowing. Mexico, the site of the breeding program that produced the high-yielding dwarf wheats that came to be widely used in the Third World, and the first developing country to achieve a yield takeoff, has become the first developing country to "hit the wall." Like the United States, Mexico hasn't seen any improvement in the past 13 years.

Rice Yield Prospects: The Role of Irrigation

To achieve its full yield potential, the rice plant requires large quantities of water, either from the natural flooding that occurs in monsoonal climates or from irrigation. When farmers are forced to rely on monsoon flooding, they lack the water control needed to effectively use fertilizer and pesticides. Irrigation, then, is a primary determinant of rice yields. Countries with high rice yields irrigate a high percentage of their riceland.

Japan, which has the highest yield among the major rice producers, irrigates 99 percent of its riceland (Table 2) - as does South Korea, which ranks second. In Bangladesh, where the irrigated share falls to 24 percent, farmers harvest less than half as much rice per hectare as the Japanese or Koreans do. Three-fourths of its riceland is subject to the whims of the monsoon.

Rice is also affected by latitude, with yields rising as distance from the equator increases. In Japan, the northernmost of the major rice-producing countries, the length of day during the summer growing season can be several hours longer than in a country like Indonesia, which lies astride the equator.

Table 2. Rice Yield Per Hectare in Key Producing Countries, 1995(1)

Country                           Metric Tons

Japan                                4.8
South Korea                          4.7
China                                4.2
Taiwan                               4.1
Indonesia                            2.9
Vietnam                              2.4
Philippines                          1.9
India                                1.9
Pakistan                             1.8
Bangladesh                           1.8

1 Yield for 1995 is the average of 1994 through 1996. Source: USDA

A third factor affecting rice yields is solar intensity. Asia, which produces 90 percent of the world's rice, is actually handicapped because the bulk of its rice is grown during the summer monsoon season when extensive cloud cover reduces the amount of sunlight reaching rice fields. Although Japan's latitudinal advantage and full use of irrigation give it the highest yields in Asia, they are still one-third below those of the U.S. state of California, where there is an abundance of sunshine.

The historical trends in rice yields in Japan, China, and India [ILLUSTRATION FOR FIGURE 3 OMITTED] provide some insight into the longer-term potential for raising world rice yields. As noted earlier, Japan tripled its per-hectare output over a century of genetic and agronomic improvements, plateauing in 1984. China made its improvements later, but it too appears to have tripled its rice yields. If the area planted to rice in China is underestimated by 20 to 30 percent, as Chinese officials now believe is likely, then the yield per hectare of rice is overstated and the actual yield may be closer to 3.5 tons per hectare than to the official 4.2. Nonetheless, if China's rice yield is adjusted for the latitudinal handicap it suffers relative to Japan, then China's farmers may be approaching the same productivity ceiling as that of Japanese farmers. That conclusion is consistent with the observation that China is using the full arsenal of yield-enhancing techniques: high-yielding varieties, nearly full irrigation, and the world's highest consumption of fertilizer (which suggests that it has little to gain by using much more).

In India, rice yields have doubled since mid-century, from roughly 1 ton per hectare to 2 tons. Production in India suffers from a lack of irrigation, and because it has shorter days during the summer growing season than Japan, it cannot realistically be expected to raise yields to anywhere near those of Japan.

Along with whatever potential may yet remain in India, the greatest remaining potential appears to lie in Bangladesh, Vietnam, and Myanmar. But for the world as a whole, the unrealized potential for raising rice yields is shrinking, suggesting that future gains will be much slower.

Perhaps the most promising source of higher rice yields is the new rice prototype being developed in the Philippines, and expected to be available for commercial use sometime after the turn of the century. Dr. Gurdev Khush, the Indian scientist who heads the rice breeding project, believes it could raise rice yields in the tropical and subtropical regions, for which it is designed, by up to 20 percent. If it were to succeed in raising yields on two thirds of the world's riceland, then it could boost the current world rice harvest of 350 million tons by an additional 50 million tons, enough to cover the growth in demand for grain from world population growth for two years. Vitally important though this potential gain is, it is a far cry from the doubling or tripling of yields that came with the first generation of high-yielding varieties.

Corn Yield Prospects: Argentina Can Help

At just over 4 tons per hectare, the 1995 world corn yield is the highest of all the cereals, well above the 2.5 tons for wheat and rice. The prospect for raising it any further rests largely with the United States and China, the countries that account for two fifths and one fifth, respectively, of the world harvest. Both have been remarkably successful to date, more than quadrupling traditional (pre-takeoff) levels. (See Table 3.)

The U.S. corn yield of 7.9 tons per hectare in 1995 is the highest of any cereal in a major producing country. The rise in U.S. corn yields started around 1940, the same time as for wheat, and for essentially the same reasons: as grain prices rose, improved varieties became available and more fertilizer was applied. After fluctuating between 1.5 and 2 tons per hectare from the Civil War to the early 1940s, corn yields crossed the 2-ton threshold in 1942 for the first time. In 1957, 15 years later, they breached the 3 ton-per-hectare threshold. Then the improvements came quickly. Five years later another ton was added, making it 4 tons per hectare. Another five years increased the yield another ton. And again. During the 15-year span from 1957 to 1972, U.S. farmers doubled their corn yields from 3 tons per hectare to 6 tons. [ILLUSTRATION FOR FIGURE 4 OMITTED]

Then the rate of growth began to slow substantially. Going from 6 tons to 7 tons took 10 years. More recently, going from 7 to 8 tons between 1982 and 1992, the rate of gain was only 1.6 percent per year. Even after these enormous gains, U.S. corn yields do not yet appear to be leveling off as wheat yields have.

Table 3. Corn Yield Per Hectare in Key Producing Countries, 1995(1)

Country                          Metric Tons

United States                       7.9
China                               4.9
Argentina                           4.3
Brazil                              2.5
South Africa                        2.4
Mexico                              2.2
Nigeria                             1.8

1 Yield for 1995 is the average of 1994 through 1996. Source: USDA

China, as a relative newcomer to modern corn production, has seen its yields rise rapidly. And while they will rise further they are not likely to reach U.S. levels, because unlike the United States, China does not reserve its best cropland for corn. Indeed, no other major corn-producing country has a region of deep, well-drained soils and near-ideal growing conditions comparable to the U.S. midwest.

Argentina, where the 1995 corn yield was 4.3 tons per hectare, may have the largest unrealized potential for raising corn yields of any country. Despite the high inherent fertility of the pampas, the country's past economic policies discouraged investment in agriculture, and little fertilizer was used. The recent elimination of heavy taxes on farm exports has set the stage for heavy investment in agriculture. That, along with high world corn prices, has already helped to double fertilizer use between 1994 and 1996.

In Brazil, where the corn yield per hectare is only one third that of the United States, farmers arc handicapped by a lack of the highly fertile, deep, well-drained soils on which corn thrives. A similar situation prevails in Mexico, where corn is grown largely by small holders on marginal land in the mountainous regions. These are typically hillside plots where low rainfall and thin soils - often heavily eroded - severely limit the yield potential.

Facing Biological Reality

For individual grains in individual countries, the historic trends show a sobering pattern. In every firming environment, where yields are increased substantially, there comes a time when the increase slows and either levels off or shows signs of doing so. It is equally revealing to look at the global trends. In doing so, we use three-year averages for the decennial or mid-decennial years in order to minimize the effects of weather variations. For example, the yield shown for 1990 is an average of the yield from 1989-91 and that for 1995 is the average for 1994-96.

During the four decades from 1950 to 1990, the world's grain farmers raised the productivity of their land by an unprecedented 2.1 percent per year, but since 1990, there has been a dramatic loss of momentum in this rise. If the former Soviet Union is excluded from the global data for 1990 to 1995, because of the uncharacteristic drop in yields associated with economic reforms and the breakup of the country into its constituent republics, then the rate of yield gain is not the 0.7 percent per year shown in Table 4 but 1.1 percent - roughly half that of the preceding 40 years. And while the first half of the 1990s is too short a period to determine a new trend, it does provide a reason for concern. In addition to the plateauing of wheat yields in the United States and Mexico, cited earlier, those in Canada and Egypt have shown no improvement so far in the 1990s.

Global trends for the three major individual grains, moreover, follow the pattern seen for grain as a whole. Rice production, which was modernized later than wheat and corn, achieved an annual increase in productivity of 2.1 percent per year between 1960 and 1990, but has dropped to 1.0 percent per year since 1990. Wheat yields grew between 1960 and 1990 at an average of 2.6 percent per year, then slowed to 0.1 percent during the 1990s. (If the former Soviet Union is excluded from the global trend after 1990, wheat yields increase by 1.0 percent.) Corn averaged 2.6 percent from 1950 to 1980, then fell to 1.3 percent in the 1980s. The rise in corn yields accelerated slightly during the first half of the 1990s, reaching 1.7 percent, largely because of a belated surge in yields in both China and Brazil.

With this slower rise in grainland productivity thus far during the 1990s, the obvious next question is whether the momentum can be regained through biotechnology. Yet, on that front too, progress is not promising. After two decades of research, biotechnologists have not yet produced a single, high-yielding variety of wheat, rice, or corn. Why haven't some of the leading seed companies put their biotechnologists to work to develop a second generation of varieties that would again double or triple yields, enabling farmers to sustain a rapid rise?

Table 4. Annual Change in World Grain Yields by Decade, 1950-95

Year      Total Grain   Rice   Wheat   Corn   Other Grains
                          (percent)

1950-60      2.0        1.4    1.7     2.6
1960-70      2.5        2.1    2.9     2.4         2.3
1970-80      1.9        1.7    2.1     2.7         0.4
1980-90      2.2        2.4    2.9     1.3         1.7
1990-95      0.7        1.0    0.1     1.7        -0.8

Source: USDA

The answer, say plant scientists, is that plant breeders using traditional techniques have largely exploited the genetic potential for increasing the share of photosynthate that goes into seed. Once this share is pushed to its limit, the remaining options tend to be relatively small, clustering around efforts to raise the plant's tolerance of various stresses, such as drought or soil salinity. The one major option left to scientists is to increase the efficiency of the process of photosynthesis itself - something that has thus far remained beyond their reach.

Once plant breeders have pushed genetic yield potential close to the physiological limit, then further advances rely on the expanded use of basic inputs such as the fertilizer and irrigation needed to realize the plant's full genetic potential, or on the fine-tuning of other agronomic practices such as the use of optimum planting densities or more effective pest controls. Beyond this, there will eventually come a point in each country, with each grain, when farmers will not be able to raise yields any further.

U.S. Department of Agriculture plant scientist Thomas R. Sinclair observes that advances in plant physiology now enable scientists to quantify crop yield potentials quite precisely. He notes that "except for a few options, which allow small increases in the yield ceiling, the physiological limit to crop yields may well have been reached under experimental conditions." This means that in those situations where farmers are using the highest yielding varieties that plant breeders can provide, and the agronomic inputs and practices needed to realize their genetic potential, there may be few options left for raising land productivity.

Viewed broadly, one can begin to see an S-shaped growth curve emerging for the historical rise in world grainland productivity. Throughout most of human history, land productivity was static. Then, beginning around 1880, Japan began to raise its rice yield per hectare in a steady, sustained fashion. By the mid-1950s, nearly all the industrial countries were expanding their grain harvest by raising grainland productivity. And by 1970, they had been joined by nearly all the leading grain producers in the developing world.

For the 15 years from 1970 to 1985, yields rose in a steady, sustained fashion in virtually all the grain-producing countries of any size. Then this unique period came to an end, as wheat yields in the United States and Mexico and rice yields in Japan leveled off. If these countries cannot restore the rise in yields and if, as now seems likely, more countries "hit the wall" in the years immediately ahead, it will further slow the rise in world grainland productivity, dropping it well below growth in the world demand for grain.

Eight Observations

Except for the general warning by biologists that grain yields would eventually plateau, there does not seem to be any record of specific warnings in the early 1980s that the long rise of rice yields in Japan, or the shorter term rise of wheat yields in the United States or Mexico, were about to level off. Nor is anyone likely to anticipate precisely when, for example, wheat yields will level off in France or China, though this could occur at any time. A review of the last half-century's experience in raising yields does, however, offer certain generalizations.

One, the slower rise in grain yields since 1990 is not the result of something peculiar to individual grains or to individual countries, but rather reflects a systemic difficulty in sustaining the gains that characterized the preceding four decades.

Two, every country that initiated a yield takeoff was able to sustain it for at least a few decades.

Three, most countries that have achieved a yield takeoff have managed at least to double, if not triple or even quadruple, their traditional grain yields. Among those that have quadrupled traditional levels are the United States and China with corn; France, the United Kingdom, and Mexico with wheat; and China with rice.

Four, once plant breeders have essentially exhausted the possibilities for raising the genetic yield potential and farmers are using the most advanced agronomic practices, including irrigation, the yield potential for any particular grain in a given country is determined largely by the physical environment of the country - most importantly by soil moisture, but also by temperature, day length, and solar intensity. These factors are fundamentally unalterable.

Five, all countries are drawing on a common backlog of unused agricultural technology that is gradually diminishing and, for some crops in some countries - such as wheat in the United States and rice in Japan - that has largely disappeared.

Six, as a general matter, the more recently a country has launched a yield takeoff, the faster its yields rise and the shorter the time between yield takeoff and level-off.

Seven, despite the slower rise in yields worldwide in recent years and the plateauing of yields in a few countries, there are still many opportunities for raising grainland productivity in most countries. These are most promising in those countries where there is room for improvement in economic policies affecting agriculture. Although most governments subsidize agriculture, some still have economic policies that discourage investment in agriculture. In these countries, the key to realizing the full genetic yield potential of crops is the restructuring of economic policies to encourage investment in agriculture, such as that now underway in Argentina.

Eight, even with a concerted worldwide effort to increase grain yields, the rise during the last half of this decade could slow still further, dropping below 1 percent per year - far below the 2.1 percent that sustained the world from 1950 to 1990.

Responding to the Challenge

This slowdown comes at a time when population growth and rising affluence are combining to drive up the demand for grain at a near-record pace. Even as the rate of world population growth edges down, the absolute number of people on the planet is projected to climb by some 80 million people per year well into the next century. Meanwhile, record numbers of people are shifting to diets rich in grain-intensive livestock products, which means that in some countries per-capita grain consumption is growing even faster than population is.

The resulting disparity between projected growth in demand and supply cannot, of course, exist in the real world. Prices will rise. That won't greatly hurt the affluent, who spend only a small portion of their income for food. But for the world's poor, particularly the 1.3 billion people who live on $1 per day or less, higher grain prices could quickly become life-threatening. Heads of households who cannot afford enough food to keep their families alive may well hold their governments responsible and take to the streets. The result could be unprecedented political instability in Third World cities.

If widespread political instability does materialize, it could affect the earnings of transnational corporations, the performance of stock markets, the earnings of pension funds, and even the stability of the international monetary system. As rising food prices threaten political stability and economic progress in a world economy more integrated than ever before, the problem of the world's poor becomes everyone's problem.

Among other things, the slower rise in world grainland productivity calls for an urgent assessment of the carrying capacity of land and water resources in countries everywhere, but particularly in the low-income countries where they are using all their available land and where the demand for water already exceeds the sustainable yield of aquifers.

In a world where land productivity is rising more slowly, the food security of the next generation depends on quickly slowing population growth. This begins with public education programs on the consequences of continuing rapid increases in population. It means filling the family planning gap by getting family planning services to those women who need them. And it means investing heavily in the education of young females in the Third World to accelerate the shift to smaller families.

Future food security also depends on a sharp increase in investment in agricultural research and in better grain storage facilities. It means boosting the efficiency of water use, largely by moving to water markets. And it means taking strong steps to protect cropland from conversion to nonfarm uses.

To mobilize on all these fronts, political leaders need to know that the rise in cropland productivity is slowing and that the growth in the grain harvest is falling behind the growth of demand. If the foregoing analysis is at all close to the mark, it would be irresponsible for the World Bank not to revise its outmoded supply and demand projections. Otherwise, its projection of surplus capacity and falling grain prices will only reinforce the prevailing complacency, and lead to potentially tragic underinvestment in both food production and population stabilization.

Lester R. Brown is president of the Worldwatch Institute.