.... 25 Years of Improvement

Part II. CGIAR-Mandated Food Crops


Production of CGIAR-mandated food crops in developing countries has considerably risen over the 25 years of CGIAR activity (Table 4 and chart 8 in vol. 2: Database). Production of rice, wheat, maize and cassava rose by up to 134 percent over the period.

Chart 14 "Food Crop Supply per Capita by Region" shows that, despite population growth, per capita supplies of wheat, rice and maize increased substantially. The declines in other crops, such as sweet potato, millet and sorghum indicate a shift toward "modern" grains partly because they benefited most from research-induced productivity gains and partly because of changing consumer preferences. The Aggregate Supply chart (13) shows a remarkable improvement by 8.5 percent in per capita supply of the aggregate of CGIAR-mandated commodities over the 25-year period. The global average, however, conceals strong regional differences. Whereas Asia and the Middle East achieved great gains in per capita supplies of CGIAR-mandated crops, Africa and Latin America showed a decline. In the case of Sub-Saharan Africa this reflects the slow growth of food and agriculture production which over many years did not keep up with population increases. In the case of Latin America the slight decline of per capita supply of CGIAR crops is due to changes in diet away from basic food crops and toward higher value foodstuffs.

The productivity gains achieved are shown in the chart 7 "Yields of CGIAR-Mandated Food Crops by Region" Over only two decades, average wheat yields doubled and rice and maize yields rose 80 percent (charts 8, 10 and 11). This is the quintessential result of a phenomenon dubbed the Green Revolution which started during the early 1960s and continues today.

Chart 8 "Developing World Production Increases Due to Increases in Yields and Area Planted" illustrates the contribution of productivity increases to total production growth, by crop. In the case of sorghum, the entire modest production gain was due to higher yields since the area harvested did not expand. In the case of bananas, yams, cassava and potato, most of the higher output resulted from expansion of area harvested, whereas the larger production of rice, wheat and maize came mainly from higher productivity. It should be noted, however, that the strong area expansion typically recorded for millet and sorghum, roots and tubers, bananas and barley tends to depress average yields because the additional land cultivated is usually of lower quality. Under the theoretical assumption of constant land quality, yields would have risen more markedly. To some extent this holds true of all crops under consideration.


The Dynamics of Developing Country Agriculture

A comparison of growth rates of agricultural production in developed and developing countries (Table 1) shows the impressive progress made by the latter group. The growth achieved by developing countries since the early 1960s is at least double (cereals) if not three and four times (livestock) that recorded for the developed world. This is particularly evident in the last quinquennium (1991/95) which shows healthy growth rates in developing countries and negative growth in all food production sectors in developed countries.

Recently, much public discussion has focused on the allegedly sluggish growth of food production in developing countries, giving rise to a renewed alarm about deteriorating levels of world food security. The FAO index numbers give little support to this concern. While it is true that growth rates in developing world cereal production have been declining, the rates continue to be strongly positive. Actually, the absolute annual increment (in quantity, not percentage) of developing country cereal output was in the last quinquennium higher than during the first phase of the Green Revolution, the 1960s. Chart 10 depicts the long term rising yield trend of the main cereals in developing countries.

The observed reduction in cereal yield growth rates does not seem to affect overall food output in developing countries, on the contrary: the latter's growth rate has been remarkable stable over the thirty year span under consideration and is now, with 3.76 percent, close to record level. Developing country food output is currently growing per year over 4 percent faster than that of the developed world. This shows, if anything, that the developing countries, as a group, are making a massive contribution toward improving world food security (Chart 3 and 4).

There is little need to discuss why the developing countries are much more dynamic in food and agriculture production than the developed world. It is evident that population growth and economic expansion are boosting demand while, on the production side, more agricultural labor input, improved technologies and better farm management expanded production. The developed countries, on the contrary, with their combination of low population growth and fast technology and management change, had to pursue policies aimed at containing food output. These policies seem to have been increasingly successful since the mid-1980s.

The general public has received little notice of the revolution in the world food sector that has taken place. The developing countries have become world leaders in most sectors of food production. In 1981/85, for instance, they surpassed the developed countries in the production of fish, crustaceans and molluscs (Chart 5) While the developed world's catches and landings are in steep decline, the developing world's production follows a long term linear growth trend, favored by the introduction of Exclusive Economic Zones (EEZ) regulating fisheries. China is now not only the world leader in rice production -- as could be expected -- but also the by far largest producer of wheat, harvesting more than the USA and Russia combined. India -- the second largest rice producer -- has in 1995, for the second time, also become the world's second largest wheat producer, overtaking the USA in both output and yields. Only in maize, the USA continues maintaining its top position. In 1978, China was the world's top producer of one crop, tobacco. Now, China is the world's largest agricultural producer and leader in grain, cotton, oilseed, fruit, and tobacco production.

Continuing and Worsening Problems

Despite these positive news there continue to be serious, and partly worsening, problems. Chart 6 contains an analysis of the production of CGIAR mandated food crops by region. The crops for which the CGIAR has a research mandate are basic food crops serving to ensure the nutrition of poor population strata. Typically, the consumption share of mandate crops is highest among the poorest in any population. With rising incomes, people and countries tend to reduce their consumption of mandate crops in favor of higher value foodstuffs -- meat, fish, dairy products, sugar and fruits.

However, a comparison between developed and developing countries (Chart 6) shows that the latter's per capita output of mandate crops is at best half that of developed countries. In 1995, for instance, developed countries harvested close to 250 kilograms of wheat per person whereas developing countries harvested only about 120 kg of rice per capita -- calculated as paddy rice which has a lower calorie content than wheat.

Overall, the chart for developed countries -- considering the different scales -- is much more impressive. Even the per capita level of barley output -- a minor grain in developed countries -- is almost twice that of wheat or maize production per capita in developing countries. In Africa, over the past 25 years, only cassava, maize and yam show some per capita growth; otherwise the chart reflects the region's recorded long term decline in per capita food production.

Chart 2 in combination with Chart 8 shows that, since the early 1960s, yields per area unit increased much faster in developing countries than in the developed world, and that these yield increases provided the bulk of the additional food crops harvested. The agricultural revolution that has made developing countries the world leader in food output was in essence and continues to be a yield revolution.

Interestingly, the rising yield trend appears to continue (Chart 10) although since the second half of the 1980s the rural population growth rate is declining in all regions except Africa. (Table 3) In Latin America the rural population has stabilized. This means that the input of farm labor cannot be increased as in the past; its contribution will gradually have to be replaced by accelerated technology and management improvements. More labor saving technologies will have to be introduced in future if satisfactory yield trends are to be maintained.

As economic wisdom would suggest, trends in developing country per capita food production appear to have been influenced by price fluctuations. However, due to scarcity of data only limited meaningful information is available on price effects. Chart 9 indicates some positive correlation between (world market) prices and developing country production trends. Only the late 1980s -- the worst years of the debt crisis -- showed that wheat and rice output did not follow the price trend, presumably because many developing countries stimulated domestic production to boost supplies in compensation of their declining food import potential. The 1993-95 surge in cereal prices is only partly reflected in this chart and is likely to further raise per capita production.

Regional disaggregation indicates a similar price/per capita output relationship. Chart 11, analyzing crops of regional significance, shows that price and output trends loosely correlate, except for the late 1980s, which -- as discussed above -- was an exceptional period.

Despite the remarkable progress achieved, most developing regions have had to supplement production with rising imports. Chart 12 shows that Africa, Asia & Pacific, and the Middle East have become heavy food importers while Latin America maintained its traditional role as a net exporter.

The CGIAR as a generator of production technology and management practices for its mandate crops is faced with two opposing trends: In Asia & Pacific, Latin America and the Middle East, the aggregate per capita supply of its mandate crops is declining since 1990, if not before. This reflects the improvement of diets resulting from higher incomes, more varied and ample food supplies, and should be understood as a sign of welcome progress.

In Africa, however, consumption of mandate crops increased and peaked again around 1990; since then it remained on a high level, indicating that impoverishment in the wake of the debt crisis and resultant food shortages forced people to return to earlier, less expensive, diets rich in mandate crops.

Environment: The Technology/Demography Gap

At the prehistoric technology level of hunting and gathering -- practiced before agriculture was invented -- the world's currently used arable land would feed only about 1 percent of the Earth's present population, i.e. some 60 million people.

The invention of agriculture and the gradual improvement of crops and livestock over 10 millennia allowed a hundredfold increase in population density by changing -- but not destroying -- the environment. Forests and savannas -- an ecologically stable system -- gave way to farms and fields embedded in another system that until very recently retained enough forests, wetlands and other marginal areas to guarantee its stability.

The introduction of farming concentrated food production on the most suitable land, saving the rest of the world's land from overexploitation by hunters and gatherers. The (comparatively) high yields offered by crop cultivation permitted land savings: Land savings are but another term for productivity gains.

Hunters and gatherers, despite their limited skills, can damage the environment when too numerous: the overexploitation of the mammoths is but one example. Farmers can also harm the environment and have done so. Massive forest losses in tropical and subtropical areas testify to that; salination of overirrigated lands, erosion of hillsides, pollution from intensive farming and livestock production are other aspects of farmers' disregarding the integrity of the environment.

The farmer tends to clash with the environment when the technology level at which she or he are operating is below effective demographic and economic demand levels. Since about 1950, technology development in the developing countries has lagged behind demographic and economic trends, resulting in environmental strain.

However, had technology not developed as it did, environmental strain would have been unimaginably stronger. Table 5 in combination with Charts 14 and 15 indicates how much land was saved from cultivation through higher yields. Had crop technology in developing countries remained at 1970/74 levels, they would additionally need the equivalent of the combined arable land of the USA, Brazil and Canada to produce today's harvests. Clearly, not even a fraction of this additional land would be available for cultivation in the developing regions -- not to mention the need for its even distribution throughout all four regions and the countries therein.

Science has helped prevent an environmental disaster of truly global proportions, and it will need to continue doing so even more vigorously in the future. The technology/demography gap must be closed if agriculture is again to become part of an environmentally stable system.

Toward a New Stable System?

Scientists essentially perform the same kind of gradual crop and farm improvement work that billions of farmers have pursued since the inception of agriculture. The difference is that scientists can improve technology much faster. They, in tandem with current and future generations of farmers in developing countries, must attain the higher levels of technology needed to protect the environment from abuse.

Currently farmed marginal lands must be allowed to return to their original state; forests must be allowed to grow back where they are essential; the biological diversity of nature must be protected and enhanced. Farming inputs must be used judiciously in order to avoid pollution by overuse and degradation by underuse, for instance of soils due to insufficient fertilization, resulting in nutrient depletion. That the developing countries are, on the whole, moving in this direction, as is apparent from Chart 18.

Mechanization, represented by the number of tractors, is decreasing in Africa, where policy change and the debt crisis since the 1980s have resulted in large numbers of tractors falling into disrepair for lack of spare parts; mechanization is still increasing in Asia and the Middle East whereas it has stabilized in Latin America. Pesticide use is sharply declining in all regions due to the progress of integrated pest management (IPM). Fertilizer consumption continues to rise in all regions except Latin America where it is slowly declining because of past price increases.

Regional Overviews of the Food Supply Situation :

Africa south of the Sahara

Cassava has remained sub-Saharan Africa's main food crop, providing between 15-16 percent of regional caloric supplies over the past twenty years (1970/74-92). Production rose 94 percent since the early 1970s, while per capita output remained remarkably stable over the period. These results were achieved by expanding the area harvested (35 percent) and increasing the yields 30 percent). Africa is the world leader in cassava output, and consumes some 80 percent of its production.

Yam output is important in Africa and has increased 88 percent over the past twenty years. Production increases were achieved by increasing the area harvested (51 percent) and increasing yields (25 percent). Per capita output has increased slightly (5.8 percent), while consumption has remained stable over a twenty year period (58-59 percent).

Maize, the most important grain in Africa, provides 15.4 percent of food supply in terms of calories, up from 14.2 percent in 1970/74. The area harvested for corn has increased 34 percent in that period, while output rose by 54 percent. Yields have increased only 15 percent, despite fertilizer inputs that equal those of most other regions. To compensate for a per capita production decline of about 13 percent and steeply rising domestic food consumption, maize imports increased sharply.

Sweet potato has remained an important crop, providing between 5-5.5 percent of regional caloric supplies over the past 20 years. Africa consumes nearly all of the sweet potato produced (88 percent in 1992). While area, production and yields have all increased (33 percent, 49 percent, and 12 percent, respectively), per capita production has decreased 16 percent due to an increasing population.

Plantains and bananas are important staples in Africa, accounting for 3.8 (1990/92) percent of calorie supply, although per capita output has declined nearly one-third since the early 1970s. The world's leading producer, plantain production rose 28 percent in Africa due to a 55 percent increase in the area under cultivation. However, plantain yields decreased 17 percent in the same twenty year period. Africa's banana production followed a similar course, with overall output increasing 25 percent. The area harvested increased 49 percent, but overall yields decreased 16 percent. The bulk of both banana and plantain production is used to meet domestic consumption needs (79-87 percent).

Sorghum and millet, taken together, exceed maize in importance, covering 15.8 percent of Africa's food supplies. Both are typical low-input dryland crops characterized by low and stagnant yields. Production and area planted in sorghum rose almost 41 and 38 percent, respectively, while millet area and output expanded 22 and 28 percent, respectively. However, yields increased only slightly (2 percent for sorghum and 5 percent for millet). The per capita production also decreased significantly for both crops (20 percent for sorghum; 28 percent for millet). Millet production used to meet domestic consumption needs has increased 14 percent between 1970/74 and 1992; sorghum production to meet these needs has decreased slightly.

Food legumes (pulses) are important in the African diet. There has been a 44 percent increase in output since the early 1970s due to an 18 percent increase in area harvested and a 22 percent increase in yields. While per capita production has decreased 18 percent over the past twenty years, the amount of production used to meet domestic consumption has remained stable at 80 percent.

Rice is a popular staple in West Africa and Madagascar. The area harvested in this crop has expanded 77 percent in the past 20 years, yields have risen 18 percent and total output rose 109 percent. This has led to a per capita output increase of 18 percent. However, these gains did not come close to meeting domestic consumption needs, and imports have risen dramatically over the past 20 years.

Wheat is becoming increasingly popular in Africa, with output up two-thirds in the past 20 years, due to rising yields (63 percent). However, as with rice, these gains did not come close to meeting domestic consumption needs and imports have risen accordingly

Potato, while contributing little to caloric needs (0.4 percent in 1992), has shown a marked increase in overall production (80 percent over the past 20 years). This increase is due almost exclusively to an increase of area brought under cultivation (71 percent); yields have increased only 5 percent in this same period. The production per capita has remained stable and nearly all potato produced goes to meet domestic consumption needs.

Asia and the Pacific

Rice is the main staple in the Asia and the Pacific region, providing almost 39 percent of calories. Asian agriculture shows a rapid increase in area, yields, and production from the early 1970s up to the late 1980s, after which all three factors leveled off. As a result, per capita rice production has been decreasing in the region since 1990. Over the past 20 years, Asia and the Pacific region have consumed nearly all the rice it produces.

Wheat is Asia's second most important staple and has been growing much faster than rice. It now makes up 19.2 percent of total calorie supply. Although the area harvested in wheat expanded only 19 percent in the past 20 years, production increased enormously (139 percent). Yields rose close to 3 percent a year, with high fertilizer inputs. Asia is the leader in all three growth parameters - area, output and yield. Wheat output is now close to half the region's rice harvests, and, as with rice, practically all wheat is used for domestic consumption. There is no leveling off of production visible; yields continue to rise and the area harvested is still expanding.

Maize is Asia's third most important grain, but with a slowing decreasing ratio of direct domestic consumption - about 60 percent is now being used as animal feed. Yields are rising almost as fast as those of wheat, averaging almost 3 percent per year, while area expansion is continuing at a reduced rate. Since the early 1970s, yields have doubled and production has risen 134 percent. Fertilizer input levels are similar to those in Latin America.

Sweet potatoes continue to be of great importance, especially in China. Asia and the Pacific is by far the largest producer, and per capita output, while decreasing 36 percent over the past twenty years, is still the highest among regions. Yields have increased only 27 percent over the past 20 years, but this was accompanied by a significant reduction in area harvested (29 percent). The share of sweet potatoes devoted to domestic food consumption has decreased by about one-third.

Cassava output almost doubled over the period. The area harvest in cassava has expanded some 44 percent, and yields rose 35 percent, making Asia and the Pacific the leader in expanding cassava production, having overtaken Latin America in the 1970s as the second largest producer. Per capita production increased 37 percent, but the part of production going to domestic food consumption has decreased 30 percent, largely because of rising exports and feed use.

Potatoes are quickly becoming a major food staple in Asia, with production up 75 percent since the early 1970s, mainly because of strong expansion of the area harvested (48 percent increase since 1970). Yields have risen 18 percent while per capita production has increased 23 percent. The part of production going to domestic food consumption has remained stable at 60 percent.

Food legumes are an important staple in Asia - although per capita production is lower than in Latin America, it is equal to that in Africa. In both area harvest and output, the Asia and the Pacific region is by far the largest producer, with more than double the combined harvests of the other two regions.

Bananas and plantains, characterized by high domestic consumption, are becoming more important in Asia and the Pacific. Banana production doubled in the 20-year period, due to a 70 percent yield increase and a 21 percent increase in area harvested. Per capita production increased 38 percent. Plantain production increased 28 percent due to a slight increase in are harvested and a significant increase in yields (19 percent). Per capita production of plantains decreased 10 percent.

Barley, while contributing only minimally to the region's caloric intake, is still a major crop in terms of area, production, and yield. Asia and the Pacific region are the largest producers in terms of area harvest and overall production, while yields rival those of the MENA region. Both production per capita and the amount of production that goes to domestic consumption have declined.

Millet and Sorghum have decreased in importance as measured by their contribution to the region's caloric intake. While both the area cultivated and output of both crops decreased dramatically, yields increased substantially. Asia and the Pacific remains the world's largest producer as measured by overall output.

Latin America and the Caribbean

Maize was the best performer, with yields rising 54 percent between 1970-74 and 1994. Output rose 62 percent and per capita production increased 5.4 percent, while the area harvested expanded only 5 percent. Maize now makes up 16.4 percent of the region's 1992 calorie supplies and continues to be the region's most important grain crop.

Wheat showed considerable growth in yield and total production between 1970-74 and 1985-89. Since 1990, the area harvested has decreased noticeably, but overall production did not decrease at the same rate, mainly because yields rose somewhat. Wheat production in the region, combined with sizable imports, accounts for almost 14 percent of food supply.

Rice is an important crops, providing 10 percent of calorie supplies. Yields rose 51 percent since the early 1970s, and production increased 55 percent. Rice provides 10 percent of calorie supplies. Per capita production remained unchanged. Only one-third of the total rice area is irrigated; fertilizer use is lowest among the major regions of the developing world. However, these gains were not sufficient to meet domestic consumption needs and, since 1990, the LAC region has had to supplement their rice production with imports.

Potatoes showed healthy increases in yield (45 percent) and total output (34 percent), while the area harvested decreased 7 percent. Per capita production decreased 13 percent over the same twenty year period. A relatively large share of potato land, 24 percent, is irrigated. LAC consumes over two-thirds of its production.

Barley production has increased 12 percent over the past 20 years, reflecting a dramatic increase in yield (49 percent), while, at the same time, the area under cultivation has decreased significantly (25 percent). However, production per capita has fallen 27 percent; since the early 1980s, the LAC region has imported barley to meet its domestic consumption needs.

Bananas and plantains are important economic as well as food crops for the LAC region. Together, they constitute 2.4 percent of the caloric needs of the region. Bananas, particularly, are an important export crop, with approximately 55 percent of production targeted for exports. In contrast, most plantain produced goes to meet domestic needs. Banana area and output have both increased over the past 20 years (36 percent and 32 percent, respectively), while yields have declined slightly. Plantain area, output and yields have all increased (30 percent, 43 percent, and 10 percent, respectively).

Food legumes remain an important crop in terms of meeting the caloric needs of the region, although their importance has decreased from 5 percent to 4 percent of overall food supply per capita in the past 20 years. The production per capita has decreased significantly (50 percent), and nearly all pulses produced are used to meet domestic consumption needs. While overall production increased 21 percent, yields rose only 5 percent, indicating production gains due mainly to an increase in area under cultivation.

Cassava and Sweet potato have both decreased in importance as a food crop for the LAC region over the past 20 years. Area harvested, overall production, yields, and production per capita have all stagnated or declined.

Middle East and North Africa (MENA)

Wheat is the main staple crop in the MENA region; MENA is also the leading per capita producer of wheat. Wheat production grew 115 percent over the 20-year period, almost as quickly as in Asia. The region experienced fast rising wheat yields, which almost doubled over the period, while the harvested area expanded only 10 percent. The MENA region consumed 160 percent of the wheat it produces, indicating the continuing need for heavy imports. Wheat covers an exceptionally high 44.3 percent of the regions total food supply.

Rice production rose 53 percent with slow area expansion and provides 6.3 percent of the calorie supply. Yields, already high because of substantial irrigation, rose an additional 34 percent to 6.6 tons per hectare, nearly twice the level of the second best performing region in the developing world, Asia and the Pacific. However, due to a rapidly expanding population, per capita production decreased 18 percent. The amount of production used to meet domestic consumption needs also increased significantly (42 percent).

Maize has developed much faster than rice over the 20-year period. Production has more than doubled, mainly due to yield increases of 68 percent, thus maintaining the region's rank in achieving the developing world's highest maize yields. Maize provides 4.9 percent of the region's calorie supply and approximately two-thirds of production goes to meet domestic consumption needs.

Potato production grew extremely fast over the 20-year period, by 240 percent. Farmers achieved a 47 percent increase in yield and more than doubled the area harvested. This growth has been accompanied by very high levels of fertilizer input, almost matching French and US levels and by high levels of irrigation. Potatoes contribute 2.2 percent (1990/92) to food supplies, up from 1.4 percent in 1970/74. Per capita production also increased dramatically (85 percent).

Food legumes, an important staple in MENA, continue to contribute 2.6 percent to the diet although per capita output has almost been halved since the early 1970s. The shortfall has been compensated by imports, permitting a doubling of the share going to domestic food consumption. While yields have increased slightly, both the area harvested and overall production have deceased.

Barley is a food/feed crop of which the region is by far the leading per capita producer. Barley provides 1.7 percent of calorie supplies, down from 2.3 percent in the early 1970s. Output rose more than 80 percent over the period, mainly due to area expansion, while yields stagnated. Per capita output remained unchanged while the already low share of domestic food consumption continued to decline in favor of more feed use.

Millet and Sorghum have decreased in importance as measured by food supply per capita over the past 20 years. The area harvest, overall production, production per capita, and amount of production used to meet domestic consumption needs have all decreased; only yields of sorghum have increased.

Bananas constitute a small, but growing food crop in the MENA region due to irrigated farming practices. In the past 20 years, overall production has increased 282 percent, due to an increase in area harvested (92 percent) and yields (100 percent). The production per capita has more than doubled; however, this increase was not enough to meet domestic demand and the MENA region continues to import bananas.



Rice is the most important food crop in the developing countries which are still producing 1.6 times as much rice (brown, dehulled) as wheat, the second most important staple. Rice now provides 29 percent of the total calorie intake of developing countries, down from 31 percent in the 1970s. Ninety-one percent of the world's rice is produced in Asia and Pacific. Studies show that the demand for rice in Asia will increase by 70 percent over the next 30 years where 2/3 of the world's poor live. Per capita output in Asia has been stagnating since the late 1980s. Latin America's and the Middle East's shares in rice output are decreasing while Africa's share is increasing. In Latin America, per capita production has remained fairly unchanged over the period; it decreased in the Middle East but is slowly increasing in Africa. Despite the dynamics of Africa's rice sector, it cannot satisfy more than a fraction of spiraling demand which has made substantial and rising imports necessary. On a modest scale, Latin America is also supplementing its domestic production by imports. Fertilizer use is lowest in Latin America and Africa because of the prevalence of dryland production.

New Developments in Rice Research

Asia and Pacific

Irrigated rice production is meeting a yield barrier that threatens future food security of the region in view of quickly growing demand. IRRI is responding to this challenge by modifying the plant architecture while maintaining the plant's level of tolerance to major pests. There is no room for expansion of irrigated riceland and water is also becoming scarce. Intensive use of the land is causing soil health problems which result in yield stagnation and a need for higher levels of inputs.

Future expansion of the rainfed riceland is limited by soil related stresses, flash floods, stagnant deepwater, salinity, drought, certain mineral deficiencies, and iron and aluminum toxicity of soils.

Pests and diseases remain a major threat to rice productivity in all ecosystems. As rice-based systems in Asia inevitably intensify in order to meet increasing demand, diseases such as sheath blight and rice blast are becoming serious problems. Pests, such as stemborers, will continue to require innovative research to preserve recent successes in integrated pest management which, when used by farmers, can substantially reduce pesticide use and diminish health risks.

The shift from transplanting to direct seeding, driven by scarcity of rural labor, and of water, will increase losses to weeds. Without new weed management tools, farms will be forced to rely more on herbicides.

Progress has been made in raising the yield potential of irrigated rice through development of the new plant type, but further research on fundamental aspects of yield enhancement such as increased carbohydrate synthesis, photosynthetic efficiency, heterosis, and nitrogen metabolism is needed. In addition, hybrid rice suitable for the tropics and which can increase yields by about 20 percent is being developed. For now, the use of hybrids requires an efficient seed sector since farmers must buy their new seed each year. However, scientists are beginning to explore, with biotechnology, the use of apomixis so that farmers can save and use their own hybrid seeds from year to year.

Fertilizer demand, especially nitrogen, will grow rapidly with the need to increase productivity in both irrigated and rainfed systems. Research to improve nutrient use efficiency is urgently needed to reduce fertilizer use. IRRI has new technologies on the horizon that promise to meet the plant's nutrient demand only at the rate required, thus reducing wastage and environmental pollution. This will open the way for resource-poor farmers, in the rainfed ecosystem, to take advantage of the benefits of fertilizer in an economically and environmentally friendly manner.

IRRI has made good progress on some abiotic stresses but needs to intensify research efforts to develop appropriate technologies. Prototypes to increase the yield potential of deeply flooded areas of Southeast Asia have been developed. Salt-tolerant, high yielding varieties, and new deepwater rice types for South Asia, are in the research pipeline.

In upland rice, weeds are the most serious biological constraint. Varieties that are highly competitive with weeds have been selected and are being tested. Some cultivars produce natural chemicals that can suppress and kill weeds in field environments. They will be used to reduce the drudgery and cost of weeding.

Latin America and the Caribbean

Once the efforts at IRRI to break the irrigated rice yield barrier are completed, CIAT will add the necessary pest resistance traits appropriate for Latin America and the Caribbean (LAC) so that LAC obtains the benefits as soon as Asia does. CIAT is also adopting a second approach to overcome the yield barrier, called "recurrent selection". This method has proven its worth in maize, delivering steady yield increases over the long term, so CIAT is adapting it to rice. Hybrid rice may increase yields another 20 percent, but seed will be expensive. CIAT is not working on hybrid rice, since numerous other institutions are. Over the longer term, apomixis may deliver the same yield increase as hybrid rice, without the high seed cost. However this may take 15-25 years.

For upland rice, which accounts for about 30 percent of LAC production, CIAT has recently achieved an exciting breakthrough in the 4t yield barrier, raising it to 6 t (50 percent). Average farm yields are expected to double, from 1.3 to 2.5 t/ha. Simultaneously, grain quality was improved in these varieties, to increase crop value by an estimated 20 percent. Enormous impact is expected, as these varieties reach the farm over the next five years.

Expansion of irrigated rice land is unlikely due to high and increasing costs of water and water control infrastructure, compared to the falling price of rice and decreasing investments in agricultural infrastructure development by LAC governments. Expansion of rainfed lowland is also questionable, because it is a technically less efficient production system than irrigated, tending to be pushed out due to competitive forces as the price of rice declines. Expansion of upland rice, on the other hand, is an exciting possibility due to new highyielding germplasm (created by CIAT in collaboration with EMBRAPA, Brazil) for the acidsoil tropical savannas. Essentially, it makes it possible to grow rice like wheat grows in Kansas, using similar machinery with low costs of production over large areas. Currently, there are 3 million hectares of this type of cultivation; it is likely to expand to 5 million hectares within the next decade, with a potential of about 15 million if market demand warrants. These varieties may also trigger the adoption of improved pasture grasses and legumes intersown into the rice (improved agropastoral systems), which would enrich soil fertility and create a more sustainable agricultural system than those presently available. The rice acts as a "nurse crop," speeding up the establishment of the pastures and paying for their establishment costs through profits from the rice sales, creating the incentive for farmers to adopt the system.

Weeds cause the greatest yield losses regionwide (estimated at 11 percent). Herbicides used to be considered the only solution, but weeds are developing resistance to current herbicides while fewer new herbicides are becoming available due to environmental concerns.

Blast fungus (Pyricularia grisea) is a major fungal disease, particularly in the tropics, with 5 percent average yield loss estimated. This average masks the uncertainty which plagues farmers: they may have no problem one year but see their crops devastated the next.

Hoja blanca virus causes destabilizing epidemics every decade or so in the tropical parts of LAC.

Sheath blight, a fungus (Rhizoctonia solani), is favored by highyield crop management practices, such as high fertilizer rates and repeated rice cropping.

For upland rice, low soil fertility is the major constraint due to the ancient, weathered acidic soils dominating the cultivated area. Native phosphorus, silicon, potassium, calcium, and magnesium are very low in these soils. Drought is another major constraint for upland rice, due to the low waterholding capacity of typical soils and erratic rainfall. Irrigation water is increasing in cost, and development of new water sources is decreasing due to costs and environmental concerns.

For irrigated rice in the Southern Cone (southern Brazil, Uruguay, Argentina, and Chile), cold is a major constraint at planting and flowering stages.

The low price of rice is a major constraint to further investments by farmers towards yield enhancement.

Tariff and quota protection of rice markets, direct and hidden subsidies, etc., by all countries inhibit progress towards maximal efficiency in production and lowest prices to consumers.

As the price of rice continues to fall globally, reducing costs of production while holding on to current yield levels will be the key factor for farmers struggling to remain in business and in turn for the rice production situation in LAC. The research challenges then will include reducing use of water and specialized machinery (e.g., through expansion of upland rice), fertilizer, herbicides (see weed discussion above), fungicides (through more durable blast resistance), and insecticides (through hoja blanca resistance). Varieties that give higher yield per unit cost of inputs, relative to the value of rice in the marketplace, will be more important to farmers than simply higher yield alone. Expansion of upland rice in the savannas, both in crop rotations (with soybean, maize, and others) and in more sustainable agropastoral systems (see above), may be an important trend, particularly in Brazil.


The primary constraint to grain yield in most African rice production systems is not inadequate yield potential of available cultivars, but rather a range of factors that contributes to large gaps between farmers' yields and potential yields achievable under improved management conditions. Among these factors are: (1) rice is a relatively new crop for many farmers, (2) conflicts in timing and quality of resources allocated to rice due to competition with nonrice crops in mixed farming systems, (3) dominance of extensive landusing rather than landsaving (yield augmenting) production systems due to surplus land, (4) diverse and unfavorable production environments dominated by rainfed conditions, (5) low use of purchased inputs due to inefficient input and product markets, and (6) large losses from weeds and insects due to low use of chemical control measures.

Driven by demographic and market forces, the region's rainfed rice farmers are attempting to increase production primarily by expanding cultivated area. With limited access to high-quality soils appropriate for rice cultivation, this is occurring increasingly through reduction in fallow periods and expansion onto marginal soils. Both processes can induce degradation of the natural resource base and declining production potential. Soil maintenance under intensified cultivation is most difficult in the uplands due to the more rapid breakdown of organic matter and to a higher rate of loss of nutrients by leaching and erosion. Acidity problems related to chemical fertilizer use is also more severe in the low activity clays of the uplands. Intensification results in reduced nitrogen supplying capacity, soil erosion, and a rapid infestation of the land with problem weed species.

The potential for sustainable area expansion is greatest in the lowlands. Lowland soils are more robust and can maintain continuous cultivation without soil degradation. It is estimated that there are 20 million ha of inland valleys in West Africa alone, of which only about 15 percent are currently cultivated. If the share of cultivated lowlands doubled and were put into rice production, this would represent a doubling of total rice area in West Africa, and still would leave some 70 percent of lowlands in their natural state to preserve biological diversity.

Across much of Africa, rice has become a commodity of strategic significance. Driven by urbanization and changing food preferences in both urban and rural areas, rice consumption in subSaharan Africa as a whole has increased more than 5 percent annually since 1970, with consumption and production spreading well beyond its traditional centers in West Africa and Madagascar. Within West Africa, revolutionary changes in consumption habits have led to a wide and growing imbalance between the demand for rice among West African consumers and the supply produced by the region's farmers. Since 1970, West African demand has grown nearly 6 percent a year, with most of this growth caused by substitution for traditional coarse grains, roots and tubers. The share of rice in cereals consumed has grown from 14 percent in 1970 to nearly 25 percent in 1990. The FAO projects that annual growth in regional rice consumption will remain high, at 4.5 percent, through the year 2000, with the share of rice in consumers' diets increasing.

Urbanization, changes in employment patterns and the subsequent broadening of rice consumption across income classes are the most important factors contributing to changing consumer preferences. Rice is no longer a luxury food but has become a major source of calories for the urban poor. The poorest urban households in West Africa obtain larger shares of their cereal-based calories from rice than do higher income households, and rice purchases represent a greater share of their total cash expenditures. In the face of rapidly growing demand, regional rice production since 1970 has risen at an annual rate of 4 percent. Although this is high compared to the performance of most other crops, at this rate regional rice production only barely exceeded population growth, and has met only two-thirds of the increments to demand. The FAO projects West African rice production to continue to grow at an annual rate of 5.1 percent.

Imports are filling the widening gap between regional supply and demand, rising at the annual rate of 6.8 percent since 1970. Rice imports, which stood at 2.2 million tons in 1980, grew to 3 million tons in 1990. The FAO projects imports to increase to 4 million tons by the year 2000 at a projected cost of more than $1 billion in foreign exchange.

Rice has also become strategically important for the region's small farmers. In many areas, rice is produced primarily by women for whom it provides an important share of their earnings. Rice is also key in strategies for sustaining the resource base. The development of appropriate lowland rice technologies is a central element in strategies to help farmers reduce pressure on degraded uplands by shifting cultivation to the lowlands.

The major challenge to rice research is to help farmers break the vicious cycle of poverty by developing low-cost rice production technologies to intensify production while sustaining the quality of the resource base. For upland rice farmers this calls for research to arrest and reverse the degradation of their fragile land base. New cropping systems, perhaps based on legumerice rotation systems, need to be developed in an integrated fertility management approach. For lowland rainfed systems, the need is to intensify production through more efficient use of residual moisture in mixed cropping systems. In irrigated systems in the humid and subhumid zones the primary need is to increase rice yields through better water management, more efficient use of purchased inputs, and through higher and more stable yielding varieties. In the Sahel, the focus of future research will be on developing short-duration cultivars tolerant to extreme temperatures to permit ricerice double cropping to increase from approximately 10 percent at present to 60 percent or more, and to improve land and water management to arrest and reverse the buildup of salinity in intensified systems. The development and transfer of integrated pest management systems is essential across all ecosystems to halt the buildup of pest problems as rice production intensifies without increased use of environmentally harmful chemicals.

WARDA scientists have recently achieved a major breakthrough by developing new plant types suited for resource poor farmers by making inter-specific crosses between O. sativa (of Asian origin) and O. glaberrima (the traditional African rice species). O. glaberrima possesses genes for resistance to drought, blast, rice yellow mottle virus, nematodes, African rice gall midge, and soil acidity, as well as for rapid seedling vigor and vegetative growth for weed suppression. The grain also has good aroma and taste. Earlier efforts to transfer useful genes between O. sativa and O. glaberrima have been frustrated by high sterility in the progeny.

Beginning in 1991, WARDA scientists began making crosses between these two generally incompatible species to create new rice plant types. Combining another culture and embryo rescue techniques with conventional backcrossing, they have substantially improved fertility rates of progeny from our interspecific crosses and have successfully introgressed genes from O. glaberrima into O. sativa. The new plant types combine the high yield potential from O. sativa with new genes from O. glaberrima for multiple stress resistance. WARDA expects that before the end of this decade new plant types developed from our wide crossing program will be extended throughout West Africa. These revolutionary varieties will meet the particular needs of resource poor farmers too often by-passed by modern agricultural research.


With an average 95.1 percent yield increase in all developing regions (1970/74 - 1990/94) wheat was the best performer among the food crops reviewed. It also was the only crop the yield increases of which closely matched the quantitative demand growth (97.3 percent) over the same period. In other words, thanks to the excellent yield increases it would have been possible to satisfy the rising developing country wheat demand without expanding acreage. However, the wheat area harvested expanded by 17 percent over the period; this indicates that wheat output exceeded quantitative demand by a considerable margin, thus bringing down prices and substituting wheat for other staples.

During the twenty years under consideration, wheat showed rising per capita production in all regions, except Africa where it slightly declined. Yields more than doubled in Asia and Pacific, the largest producer by far, and in the Middle East. In all regions, except Latin America, wheat showed the highest yield growth among grains. In terms of importance in total developing country production, aggregate wheat output rose from roughly one-third of aggregate rice output in the early 1970s to half of rice production now. Wheat now provides one-fifth (1992) of total developing country food supply, up from 15 percent in the early 1970s.

The Middle East is about to overtake Latin America as the second largest wheat producer among developing regions. However, despite high growth rates of output and yields, all regions, except Asia, are supplementing their domestic production with imports, with the fastest import growth recorded in Africa. Fertilizer inputs are high in Asia, the Middle East and parts of Africa's small production but lower in Latin America.

New Developments in Wheat Research

Over the past 30 years the spread of improved wheat varieties throughout the developing world has been phenomenal. Two CGIAR centers, CIMMYT and ICARDA, have contributed to developing and distributing improved varieties in collaboration with national research institutions.

Future progress in international wheat breeding-measured in terms of farm level impact-requires meeting at least two challenges. The first is that improved wheats, nearly all of which are related to ClMMYT's strategic research, have already spread over most of the environments amenable to production. This means that, for the area where improved varieties can expand, research must increasingly address the more difficult production constraints associated with marginal environments. And that portends slower progress on the research front. The CIMMYT/ICARDA program in Syria, for instance, is addressing these challenges by capitalizing on the genetic potential of landraces and wild relatives of wheat indigenous to the Middle East.

The second challenge is that most farmers in favored environments, who experienced the first wave of the wheat revolution, have already replaced their Green Revolution varieties with second and often third generation materials. New materials for these better environments then, whether produced by CGIAR or by others, must now compete on a much higher plane for yield, disease resistance, and a host of other factors. That means CGIAR research will have to be more innovative to achieve significant gains at the margin.

So what makes CIMMYT optimistic about future progress? For one thing, during the past decade CIMMYT wheats have spread over some 16 million additional hectares-nearly all of that rainfed and more marginal for wheat production. A considerable portion of research work is aimed at making the wheat plant more efficient in using available moisture, nutrients, and sunlight and CIMMYT has solid evidence of success. As but one example of improved efficiency, experiments done in 1993 in Mexico, clearly demonstrated that newer wheats are much better at extracting nitrogen from the soil and converting them to grain. The newer materials yielded more than older varieties at all levels of nitrogen application, even when none was applied to the soil. The implication is that these wheats will prove useful to those who farm less fertile soils, and also that when additional nitrogen is available, more will be absorbed for grain production and less leached into ground water supplies.

Such progress in wheat breeding rests on continual, long-term efforts to broaden the crop's genetic base.

In another push to broaden the genetic base of CIMMYT wheats, the scientists began in the 1980s to incorporate highly productive Chinese winter and spring wheat germplasm into CIMMYT's general populations. This work provided another 10 percent upward shift in yield potential, which will find its expression in farmers' fields during the next five to 10 years, especially in high rainfall areas.

And more recently, CIMMYT developed experimental lines that produce more grainbearing spikes per unit area, boosting the yield potential still another 5-10 percent. The best of these lines, "Attila" was developed by incorporating US and European wheat germplasm into CIMMYT materials. Because of the lags involved, however, farmlevel payoffs to this work will only become evident some 10-15 years from now.

Finally, for warmer environments, CIMMYT and ICARDA developed wheats from intergeneric crosses with wild, distantly related species that are resistant to hellminthosporium spot blotch, a disease that has effectively blocked the spread of spring wheat into such areas. These materials are now incorporated into CIMMYT's mainstream breeding program and are in the hands of national program researchers in countries where spot blotch is a significant constraint to wheat production. Impacts at the farm level will likely become evident about the year 2000.


Maize is the developing world's third most important food grain, providing 7.8 percent (1992) of developing country food supplies (1970-74: 7.9 percent). In all regions except Latin America, the maize area harvested expanded by between 20-34 percent over the period. Yields showed strong increases of between 1.7-2.7 percent per year in all regions except Africa, with Asia in the lead. Due to the combination of area expansion and yield growth, production grew fastest in Asia - 134 percent. Asia now produces more than double the maize harvest of Latin America, the second largest producer where output rose 62 percent over 20 years. Africa achieved an output growth of 54 percent over the period. In the Middle East, production more than doubled, mainly due to the high rate of yield increase. The yield growth mirrored the level of fertilizer inputs, with the Middle East using twice as much fertilizer per hectare than Asia and Latin America. Africa's relatively high fertilizer input is a statistical artifact caused by Zimbabwe's dominating role among the countries representing the region.

The share of domestic consumption has risen strongly in Africa, indicating that food demand for maize is rising faster than production, culminating in substantial imports during the peak drought year of 1992. The food consumption ratio remained steady at a low level in Latin America and decreased in Asia and the Pacific, and the Middle East, due to higher feed use.

New Developments in Maize Research

With increased pressure on natural resources, new maize varieties and farming techniques are needed that are specifically tailored to more marginal environments. Breakthroughs in this work are difficult to achieve, partly because of the greater variability in soil quality in marginal environments. After years of research, CIMMYT has developed promising selection methods, a pool of maize that can serve as a source of low-nitrogen tolerance, and even experimental varieties that produce more grain in both fertile and poor soils.

CIMMYT breeders are developing drought-tolerant maize varieties to avoid losses such as recorded in southern Africa in 1991-92 when drought destroyed two-thirds of the maize harvest. One-half of the 60 million hectares planted in maize in the developing countries is subject to periodic droughts. CIMMYT expects that as many as 20 countries in the next 10 years can begin full-scale production of the drought-tolerant maize.

Some 22 million hectares of maize land in developing countries consists of acid soils. Aluminum toxicity in acid soils is particularly harmful to maize. CIMMYT researchers over the past 10 years have selected several hundred varieties for evaluation under a range of acid soil conditions. The roots of acid-tolerant maize plants exude more citric acid than other plants, thus binding the aluminum. Six new acid-tolerant breeds also increase yields on non-acid land, and increase yields on acid soil by 40 percent. Varieties have been sent out for testing.


In sub-Saharan Africa there is no ceiling visible for maize yields. Productivity is generally lower in the humid zones with frequent cloud cover than in the moist savanna. Key constraints to area expansion are rusts, blights and streak virus which curtail production in the moist savanna zones. IITA has now developed varieties resistant to these diseases with the result that the maize area is increasing dramatically in the moist savanna. However, stem and cob borers, striga weed, streak virus and downy mildew still cause losses between 30 to more than 50 percent. Once harvested, maize is attacked by a highly destructive storage pest, the larger grain borer.

IITA succeeded in solving the maize streak virus problem by developing a large number of resistant varieties. Recently, IITA also developed and distributed varieties resistant to the downy mildew. Considerable progress has been made in combating stem and cob borers, based on a combination of biological control and resistant varieties that is in the research pipeline. Varieties resistant to the parasitic striga weed are object of an ongoing project, as well as farming practices that would remove the striga constraint to maize production.

For the future, intensified maize production in Africa will require more research on improved agronomic practices and find environmentally beneficial solutions to increased pest, disease and weed pressure.


Millet and sorghum jointly contribute 4 percent to developing country food supply (1992), down from 6.7 percent in the early 1970s. Gradually, Africa is becoming the main producer of millet and sorghum. African sorghum harvests increased 38 percent and millet output 28 percent over the period whereas other regions registered stagnation or partly substantial production declines. However, Asia is still the leading producer, especially of millet. In Africa, the percentage of millet used for domestic food consumption is rising steadily; the corresponding share of sorghum remains steady on a high level. In the other regions, the ratios of consumption show a mixed pattern. In Africa, the vast and still expanding millet and sorghum areas are producing low and steady yields with negligible fertilizer inputs. While millet yields are not much better in Asia and the Middle East, sorghum yields there are double the African level, and Latin America achieves the highest millet and sorghum yields. In sorghum production, Latin America's fertilizer input is high but so is the Middle East's-a statistical artifact because fertilizer data are available for Egypt only. While food consumption ratios of millet and sorghum are high in Africa, Asia and the Middle East, sorghum in Latin America is mainly used as animal feed.

New Developments in Millet and Sorghum Research Millet

The main constraints to increasing productivity of pearl millet that could be overcome by research are water deficit/drought, deficits in soil nutrients, and excessive heat. While crop improvement research can partially help increase drought and heat tolerance, over four-fifths of the potential yield gains from overcoming these constraints must come from improved natural resource management, says ICRISAT. However, crop improvement can help much to control current millet pest, disease and weed problems such as downy mildew, head caterpillars, stem borer, striga and other weeds. The total annual millet yield loss from abiotic and biotic stress that could be overcome by research has been estimated by ICRISAT at over $1.2 billion.

Stem borer in the African Sahel is causing annual losses estimated at $100 million. Destruction of infested crop residues, application of sex pheromones to disturb reproduction of the pest, and locally-made cheap traps are successful control measures recommended by ICRISAT.

Resistance to downy mildew, identified by ICRISAT in germplasm from India and Africa, is being transferred to economically important millet hybrids and open pollinated varieties, increasing staple food productivity of millions of inhabitants of the harshest environments in Asia and Africa.


As regards abiotic constraints, drought and water deficits, as well as deficits in soil nutrients, and occasionally water excess, have been identified by ICRISAT as the main causes of current sorghum yield losses-estimated at $1.4 billion annually-that could be eliminated by research. Drought tolerance can be enhanced by crop improvement research, but by far the highest potential yield gains could be achieved through resource management research, says ICRISAT. The losses caused by biotic constraints and weeds that could be tackled mainly by crop improvement research also total about $1.4 billion a year, according to ICRISAT.

Among the numerous sorghum pests and diseases, sorghum midge is a very destructive pest causing annual losses of more than $300 million in India, Yemen and Africa. An ICRISAT-developed resistant variety yields 50-100 percent more than commercial varieties where midge is endemic. Avoiding the use of insecticides contributes to environmental sustainability. Within the next two to three years, midge resistance will also be transferred into hybrid sorghums.

Sorghum is an important staple in the cool tropical highlands of Eastern Africa, Lesotho, Yemen, and Mexico. ICRISAT is working on improved cold-tolerant cultivars, based on material developed in Latin America, that have the potential to at least double yield. Locally adapted cultivars will be available to farmers by 1998. A farmer adoption rate of 45 percent is expected. Grain quality and resistance to leaf diseases would be superior in these new cultivars.


Barley is an important staple in Asia and the Middle East. In the Middle East it is the most important feed source for small ruminants. Asia is the leading region, producing almost two-thirds of all developing country barley. Both the Middle East, with roughly half of Asia's output, and Asia strongly increased their barley production over the period, with the Middle East being the growth leader. While Middle Eastern yields stagnated, yields rose in the other regions. While most barley in Latin America and Africa is for domestic food consumption (including processing), in the Middle East and Asia about 70 percent is used as animal feed. Barley's contribution to developing country food supply has decreased from 1.1 percent in the early 1970s to 0.7 percent in the early 1990s.

New Developments in Barley Research

Barley is a mandate crop of the CGIAR only in the Middle East and North Africa region, and in Latin America. It is a hardy, drought and cold-resistant grain. Cold early in the season and drought and heat during the reproductive stage are key abiotic constraints to barley production. Several diseases affect barley in the Middle East and North Africa, but fungal diseases are by far the most important. Some fungal diseases, such as rust, can assume epidemic proportions. Resistant cultivars which tolerate rust one year may become susceptible by the next epidemic. The epidemics of yellow rust during the last ten years wiped out widely grown, high yielding varieties of barley and wheat. Nematodes in barley may aggravate the effect of droughts. In China, an estimated 0.2 million hectares are now planted with disease-resistant varieties.

ICARDA succeeded in developing and releasing varieties resistant to most major diseases. Barley cultivars that maintain their height under drought have been released in Syria. Varieties adapted to low-yield environments have been released in various countries. Land-race-based cultivars with improved disease resistance and straw quality (for feeding purposes) have been released in Middle East and North African countries. Genotypes with tolerance to excess soil boron have been identified, giving 50 percent higher yield than traditional cultivars. ICARDA has developed cold-tolerant varieties giving 20 percent higher yields in the high plateau areas. The use of phosphat fertilizer has been shown to boost barley yields substantially in the prevailing phosphorus-deficient soils, even in seriously water-limited environments. To improve the use of barley as human food, ICARDA has developed hull-less high yielding varieties for faster food preparation.

CASSAVA (Manioc)

The starchy cassava root is Africa's main staple, with that region accounting for almost half of the developing world's output. Both in Africa and Asia and the Pacific-the main producer regions- cassava output rose fast; yields and area harvested increased fairly steadily. Since 1990, some plateauing is now visible in both Asia and Africa. While the share of domestic food consumption decreased in Asia, reflecting higher exports and feed use, the ratio increased slightly in Africa, indicating that food demand is rising somewhat faster than output. Latin America's output, yields and area harvested stagnated. High fertilizer consumption per hectare in Asia corresponds with high yields whereas Africa's lower yields match the region's lower fertilizer input. Latin America's main producer, Brazil, however, achieves high yields with very low fertilizer input.

New Developments in Cassava Research

Globally, CIAT estimates the cassava yield potential at 23.2 tons per hectare (fresh weight basis, i.e., 6065 percent moisture content), which is 126 percent above the current reported world average yield of 10.2 t/ha.

Breeding is making substantial progress in elevating the yield potential. The yield potential is related to the agro-ecosystem under cultivation, and CIAT has been stratifying its breeding work according to those zones. Yields (of dry matter) have been increased, relative to the best local and/or released cassava varieties, by 45 percent for the subhumid lowland tropics, 40 percent for the acid soil savannas, 70 percent for the midaltitude tropics, 90 percent for the highland tropics, 12 percent for the subtropics, and 46 percent for the semiarid tropics.

In Africa, cassava is a popular food in the humid equatorial lowlands, but expansion northward towards the Sahel is limited by low rainfall. Cold temperatures in higher elevations above 2000 meters (6,600 ft) are a limiting factor: mountainous areas in Africa and Latin America are too cool to grow cassava.

In Latin America, demand for cassava as a human food is stagnating, The root is replaced by other staples in urban diets. Diversification of enduses (commercial, animal feed, starch, etc.) can overcome this constraint in the near future.

A constraint to production is rapid post-harvest deterioration of fresh cassava, requiring speed and incurring high costs of preservation and transportation to urban markets. The lack of assured marketability discourages increases in planting area. Processing technologies need to be adopted in rural areas to overcome this constraint and markets developed for the processed products.

Biotic constraints, including all cassava diseases and pests, account for 31.7 percent or 3.2 t/ha loss of yield potential (not including post-harvest losses). On a percontinent basis, biotic stresses account for 49 percent reduction of yield potential in Africa, 28 percent in Latin America, and 5 percent in Asia. The high importance of pests and diseases in Africa is mainly caused by African common mosaic virus and cassava green mite. In contrast, biotic constraints are minimal in Asia.

Low genetic yield potential, poor quality of planting material, and weeds are also major biotic constraints. Globally, these three constraints cost an estimated 44.7 percent of yield potential. Low yield potential and weed losses combined do not differ much across continents, but the weed constraint is much more severe in Africa and Latin America than in Asia.

Soil constraints, including low soil fertility, soil erosion, salinity, and soil temperature, account for almost 30 percent of the foregone yield potential. On a percontinent basis, soil constraints account for 27 percent in Africa, 28 percent in Latin America, and 35 percent in Asia. The high importance for Asia represents the dramatic soil fertility and erosion problems of upland cassava systems on that continent.

Climaterelated constraints, including drought, waterlogging, and cool temperatures, account for almost 13 percent loss in yield potential. By continent the figures are 16 percent for Africa, 7 percent for Latin America, and 11 percent in Asia. The relatively large value for Africa is due to drought in semiarid areas. The current yield in semiarid areas of Africa is estimated at only 6 t/ha, 3 t below the continent's average yield.

Suboptimal management practices constrain yields by another 7.4 percent on a global basis. There are no large differences between continents for this parameter.

The slow multiplication rate of cassava planting material (cassava is multiplied through vegetative propagation of stem cuttings, as opposed to more easily multiplied seed crops) and the virtual absence of organized multiplication systems slow the diffusion of improved varieties.


The growing economies of Southeast Asia provide good growth prospects for cassava starch and starch-derived products. The opening up of the Chinese and Vietnamese economies offer particularly good opportunities for growth in cassava production. In Thailand, cassava production will stabilize as a greater proportion of production switches from pellets to cassava starch and to a lesser extent from export markets to the domestic market.

Latin America and Caribbean

Cassava production in this region has stabilized after a decline in the 1970s and early 1980s. Prospects for growth depend on product and market diversification. This in turn will depend on reducing government policy biases towards competing carbohydrate sources, principally cereals. Current trade liberalization trends and reductions in subsidies suggest movement in this direction.

Research challenges:


Cassava production is growing fastest in Africa. Cassava's dominant role as a household food security crop will be maintained, especially in those areas prone to periodic drought, and the trend of increasing offfarm sales will continue as urbanization accelerates.

Research achievements:

IITA has discovered spontaneous polyploids (varieties with multiple sets of chromosomes, e.g. triploids and tetraploids) in cassava, characterized by enormous vigor and variation in form and structure. Selections from the triploids ("super cassava") have doubled the yields of existing improved varieties with normal chromosome numbers. Introduction of earlier high yielding IITA cassava varieties has already greatly increased Africa's cassava yields and output, according to FAO. IITA breeders are busy incorporating other desirable features (disease and insect resistance, food quality) in the polyploids. When released and widely adopted, the polyploids would be a major breakthrough in the quest for better food security in Africa.

Other research achievements: The introduction to Africa, through IITA, of a wider genetic base for cassava improvement, focusing on materials with resistance to mites, mealybugs, cassava bacterial blight, tolerance to drought, low cyanogen potential, and good cooking quality. Identification, evaluation, and shipment of natural enemies for the control of the cassava green mite and the cassava mealybug. IITA and CIAT launched the famous biological control campaign of the mealybug and initiated a regional post-harvest research program for Eastern and Southern Africa.

IITA is currently testing varieties mainly derived from genetic material from the dry northern Brazil region for better drought tolerance. These would enable the crop to spread cassava cultivation into the dry savanna and parts of the semi-arid zones, thus improving considerably Africa's food security situation.

Research challenges:

Future research objectives are: resistance and integrated management of African Common Mosaic Virus and cassava green mite; adapted germplasm for drier areas and mid-elevation areas. Also needed is strengthening national research systems, facilitating germplasm flow among and within countries, and multiplication of improved planting material.

Opportunities for product and market diversification are excellent in several countries, such as Nigeria, Uganda, and Malawi. The key factor for the future of the crop will be the investment in appropriate means for the validation, adoption, and diffusion of production and processing technologies.


Sweetpotato is essentially a traditional Asian and African crop, more important than potato in terms of production and calorie content. In Asia and the Pacific, steadily rising yields permit reducing the area planted while still harvesting high but gradually declining levels of output. In Africa, area harvested and production have risen by about 33 and 49 percent, respectively, over the period, with stagnating yields and a high share of output devoted to domestic consumption. Sweetpotato which in the early 1970s contributed a strong 5 percent to developing countries' food supply, is now down to 2.2 percent, due to declining or stagnating per capita production in all regions.

Potato output is rising quickly in all developing regions, especially in the Middle East where production increased by over 240 percent since the early 1970s. Asia and the Pacific, which registered a 75 percent growth in production, is the largest producer. In Africa, starting at a low level, potato harvests increased 71 percent whereas Latin America's output rose by one-third. Latin America and the Middle East achieved yield growth of about 45 percent; Asia's yields grew 18 percent while Africa's yields stagnated. Not surprisingly, the Middle East and Latin America are leading in terms of fertilizer input, followed by Asia and the Pacific. Despite the output growth, potatoes are still contributing only 1 percent to developing countries' food supply (1970/74 - 1992), although the share of production devoted to domestic human consumption continues to be high.

New Developments in Potato and Sweetpotato Research


Potato yield potentials are well over 100 tons per hectare, the equivalent of 25 percent of grain. There is considerable room for yield increases using existing genetic materials. Many developing country farmers could easily double yields of varieties adapted to local growing conditions simply by accessing clean (virus-free) sources of seed. The same is true for sweetpotato. Technologies needed to address this problem are now available.

The main climatic constraint limiting area expansion for potato is heat. Potato responds positively to cooler temperatures. Conversely, sweetpotato favors hotter, warmer climates; production tapers off as temperatures fall. That said, there is excellent potential for expanding production area of both crops. For example, potato is now widely grown between rice or wheat crops during fallow periods in Asia.

The main biotic constraints for potato are late blight, bacterial wilt disease, and potato tuber moth. It is estimated that developing country farmers spend $700 million annually to control these pests. Potato's susceptibility to these pests and diseases make it the number two consumer of agricultural pesticides worldwide, following cotton. Sweetpotato is affected by weevils and virus diseases, which in certain regions pose major threats to production. In addition to temperature extremes, potato is susceptible to abiotic stresses of drought and salinity; sweetpotato has outstanding drought and salinity tolerance.

Previous CIP research and extension efforts using late blight resistant varieties in Africa provided an annual internal rate of return exceeding 90 percent. The Center's investment of just $5.6 million in East and Central Africa is providing annual rates of return of $10 million. In China, the use of a moderately late blight resistant cultivar provided through CIP has led to major production increases. Chinese authorities estimate that cultivar CIP24, an Argentinean variety selected at CIP for use in China, is now grown on 250,000 hectares.

In Viet Nam, CIP has helped to build a highly effective national seed potato program. Economic studies based on Viet Nam's experience indicate that the rate of return on investment in rapid multiplication techniques exceeds 100 percent.

Over the past two decades, CIP has led a global effort to develop practices that control pests without chemicals. Among these are biologicallysafe practices such as friendly fungi and natural predators, as well as potato cultivars with natural resistance to pests. Among the latter is a potato plant with hairy leaves that traps and kills insects. When combined in package form according to local farming conditions, a process known as integrated pest management (IPM), these technologies can cut back or eliminate most insecticide sprays. Prior experience in North Africa indicates that such effort provides an annual internal rate of return of 67 percent, a figure which does not include numerous benefits to the environment and human health. In Tunisia, the net benefits from previous IPM work exceeded the annual cost of the entire national potato research program.

In recent years the stability of potato crops in the Andes, the center of origin of the potato, has been threatened by infestation of Andean potato weevils. Weevils routinely damage 50 percent of a farmer's crop. The problem is further complicated by the widespread use of hazardous and largely ineffective insecticide sprays. In response to this problem, CIP has developed a series of practices that control weevils without chemicals.

CIP economists are confident that efforts to address new threats from late blight disease will provide significant returns in the future. Adoption of improved varieties now in hand should lead to production increases of at least 2 to 5 t/ha-equivalent to $400-1,000 per hectare-on approximately 2 million hectares worldwide.

In the coming years, potato production in South Asia should receive a major boost from a seed production system known as true potato seed or TPS. The use of true seed (seed produced by the flowers of the plants) will replace an estimated 20 percent of the seed tubers (seed derived from potato tubers) in marginal areas, such as rainfed mountain production zones. CIP scientists believe that by the year 2020 half of India's potato production will be derived from true seed hybrids.

Since the early 1960s, potato production has expanded dramatically in developing countries. Over the next decade, a good proportion of the increase in potato production will be on irrigated land that is also used for cereal crops. The potato fits well into farming systems with highyielding, short duration rice, wheat and maize.


Recent trends in sweetpotato production and the area planted to the crop have been highly uneven. Sweetpotato output expanded rapidly in Kenya, Rwanda, Burundi, North Korea, and Madagascar because of rapid population growth, resulting in more pressure on farmland for food. In most other countries, however, sweetpotato production has not kept pace with population growth. Declining area, which appears to be mainly responsible for reduced production, is symptomatic of weak commercial demand for fresh roots. Although considered a "poor man's food" or a survival crop in many countries, it is also eaten as a seasonal vegetable and, under certain market conditions, can actually be sold at a higher price than potatoes.


Yam is essentially an African crop with some production also in Latin America and Asia and Pacific. Africa is the leader in all performance parameters. Yields increased 24 percent, the area harvested expanded by half and output rose 85 percent. The share of domestic food consumption remained unchanged at close to 65 percent. In Latin America and Asia and Pacific production rose 31 and 34 percent, respectively. While domestic consumption in Asia and Pacific remained steady at a high 76 percent, it declined in Latin America to 65 percent, reflecting increased feed use.

New Developments in Yam Research

Yam is best suited for the moist savanna zone in Africa. In the more humid zones it suffers heavily from various pests. A large gap exists between current farmer yields and the potential yields obtained at research stations. Yam is still very much an under-exploited crop. Yield losses due to biotic stress vary according to yam species and location. Anthracnose, caused by a soil-borne pathogen, for instance, causes losses from 20-80 percent. Nematodes and yam rot caused by a complex of fungi and bacteria are other important loss factors.

For biological reasons-yam poorly responds to conventional breeding techniques-it was for a long time very difficult to improve the crop. IITA scientists, however, after more than two decades of research work, succeeded in improving the flowering qualities which can be manipulated by crop breeders to generate new varieties. With the development of the new genotypes of yams through seed hybridization, new materials are being made available both in seed and tissue culture to a number of African countries. This should help overcome some of the major production constraints such as relatively low yield and losses caused by pest organisms. These advances should allow for meeting the current high demand for yam at more affordable prices.


Bananas and plaintains together continue to contribute about 1 percent to the developing countries' food supply (1970/74-1992). While Latin America is the world leader in banana production for export, Asia and Pacific is since 1990 the world's largest producer and has a strong lead in output for domestic consumption. High yields based on high fertilizer inputs characterize the small but highly dynamic Middle East production, supplemented by imports. Asia and Pacific more than doubled its output over 20 years. Yield increases accounted for 70 percent of production gains. Latin America's output grew by a third and Africa's by a quarter but both regions registered decreasing yields. Although Africa's banana acreage now equals that of Latin America and Asia and Pacific, output is only about one-third because of low yields.

Plantains are mainly produced in Africa where they are more popular than bananas and constitute a fast growing foodcrop in terms of area harvested and output, while yields stagnate at a low level. The share of Africa's domestic consumption is rising steadily, reflecting a demand that is rising faster than output. Latin America's plantain production and area harvested are also steadily increasing, with a high share of domestic consumption.

New Developments in Plantain Research

Both plantain and banana require a lot of water and, therefore, produce well only in high rainfall areas.. Since plantain is a major basic food in Africa, improving yields is of great importance for future food security. Research station yields are four times the current farmer yields in Africa. Black sigatoka disease (named after a valley in Fiji), caused by a fungus, is the greatest threat worldwide to plantain production. Other major biotic stresses are banana weevil, cosmopolites sordidus, nematodes, and fusarium wilt. To date, most research work has focused on the development of hybrids resistant to black sigatoka. INIBAP has already distributed black sigatoka resistant hybrids from its germplasm collection. In 1994, IITA received a research award for developing hybrid plantains resistant to black sigatoka, which causes yield reductions of 30-50 percent. Even so, Africa's plantain output is said to exceed in gross value the region's maize, rice and cassava harvests.

The development of the resistant tetraploid hybrids and better understanding of the genetics of this crop offer new opportunities to increase the yield potential, better adaptation to different conditions, and improving resistance to major pests and diseases. Future IITA research will also focus on improving post harvest utilization.

INIBAP, under the governance of IPGRI, has within the CGIAR the global mandate on banana and plantain and is stimulating research and breeding programs worldwide. It coordinated the testing of new hybrids at 33 sites in all major producing regions. INIBAP organized a network encouraging cooperation among breeders and rationalizing breeding work. It is also active collecting and characterizing banana and plantain germplasm.


Food legumes (pulses) are an essential staple because of their high protein content, especially in diets low in animal protein. Overall, however, this food commodity has shown sluggish development. Per capita production declined in all regions except Asia & Pacific, and declined fastest in the Middle East where imports covered the supply shortfalls. Africa is leader in terms of yield growth (22%) but Asia & Pacific has greater area and production of food legumes. The contribution of pulses to the developing countries' food supply decreased from a high of 4 percent in the early 1970s to 2.7 percent in 1992.

Asia is by far the largest producer; Africa has overtaken Latin America as the second region in terms of output. Yields stagnated in all regions. Because of continuing rapid area expansion, Africa is-despite current population growth-maintaining its high per capita output close to Latin America's lead position. Fertilizer use in food legume production is low in all regions.


Yields of common bean (Phaseolus vulgaris) have been difficult to increase. The bean plant is already quite efficient, with a harvest index in the range of 0.550.60 (share of beans in overall plant biomass). This is comparable to the most efficient modern cultivars of rice, wheat, and maize. The plant appears to be sourcelimited (i.e., increased photosynthesis and/or decreased respiration rates would be necessary to increase yield). Different bean types have different yield potentials; later maturity and indeterminate plant type have been successfully used to increase yield (up to 6 t/ha). However, changing such adaptive traits is difficult for many farming situations. Novel genes and gene combinations are being sought which could increase yield within particular bean types through intergene pool and interracial crosses.

Greater heat tolerance and resistance to diseases, such as web blight, are needed for bean area expansion into the tropical lowland regions of Latin America. In mid to highaltitude regions, beans have moved into marginal areas, but yields are low because of drought and low soil fertility. In more favorable environments and large farms, mechanically harvestable bean types are needed to compete with soybeans and other grain crops.

An estimated 60 percent of bean production in developing countries suffers from low soil phosphorus availability. Fertilizer is often not available or too expensive for poor farmers. Besides the direct and drastic effect of P deficiency itself, it also appears to be the main factor limiting biological nitrogen fixation. N fixation in common bean is lower than in many other grain legumes. Poor fixation is primarily caused by environmental constraints. Such constraints cannot be overcome by inoculation with improved Rhizobium strains.

Expansion of bean production areas is largely limited by poor tolerance of beans to high temperature. About 70 percent of bean production in Latin America suffers from moisture deficit. The situation in Africa appears similar.

Research challenges:

For farmers, beans are an attractive crop because of its adaptability to different cropping systems and short growing cycle. The disadvantage of the bean plant has been its susceptibility to many diseases and climatic stresses. However, as greater yield stability is achieved through appropriate combinations of disease, insect, and abiotic stress resistance genes in new varieties, bean production may be an increasingly attractive option for tropical cropping systems.

Tropical snap beans (fresh immature bean pods for human consumption): There is an increasing demand for vegetables in Latin America, Africa, and Asia for consumption and export. Tropically adapted snap beans offer poor farmers opportunities to increase their income. New varieties tailored for small farming needs are becoming more readily available through breeding efforts at CIAT.

Latin America: Smallfarm cropping systems will continue to predominate in much of Latin America in the future. However, in certain areas of Mexico, Brazil, Argentina, and Chile, increased bean mechanization will be needed to satisfy urban markets demands and for export purposes. Bean consumption is expected to increase by 1.7 percent per year. Potential for area expansion is minimal unless new sources of heat tolerance become available. IPM systems will be needed to control growing pesticide abuse on snap and dry beans. Future bean varieties will contain genes for multiple disease and stress resistance.

Africa: Although total bean production in Africa is less than that of Latin America, future demand will be nearly double that quantity. The potential for area expansion is limited and most production increases must come from increased productivity. Improved germplasm and greater use of selected inputs are needed. The highest yield losses come from uneven rainfall distribution and low soil fertility. New bean cultivars with greater tolerance to both drought and low soil phosphorus will provide greater yield stability.


Chickpea is an important foodcrop in Asia, the Middle East and North Africa. Water deficits and soil-related problems, as well as pests and diseases, cause some $2.5 billion annual losses in chickpea production, according to ICRISAT. A joint ICRISAT/ICARDA program has developed high yielding chickpea lines resistant to the ascochyta pathogen which causes losses up to 40 percent. Some 50 new cultivars released in 19 countries have helped stabilize production and made winter sowing possible in the Mediterranean region. Winter sown crops yield 50-100 percent more than spring sown crops. Adoption in potential areas would result in $500 million annual benefits to farmers. Winter sowing has also enabled mechanical harvesting. Root rot resistant, large-seeded materials have been released for commercial production in North Africa.

ICRISAT has developed shorter duration cultivars that better exploit stored soil moisture; they are now being introduced in India. Work on genotypes with better drought resistance is going on; within ten years' time the results are expected to be made available to national research systems. ICRISAT is also working to improve cold tolerance in chickpea and combine it with other desirable traits. Earlier maturing cold-tolerant chickpea cultivars are expected to spread in target areas by 2005. Wild species of chickpea, native to the Middle East, are extensively used in breeding programs because of their resistance to multiple disease.


Losses in groundnut production that could be avoided through research are estimated at about $3 billion annually, according to ICRISAT. Water and soil nutrient deficits, weeds, a variety of pests and diseases, and low-yield potential of currently grown varieties are seen as the major loss factors.

Rosette disease alone causes annual losses in small farmer production in Africa estimated at $100 million. ICRISAT research identified resistance sources in genetic material used in a hybridization program. Evaluation took place in southern Africa in 1994. Cultivars ready for farmer use should be available by 1996/97, with the impact becoming visible by 2005.

Since 70 percent of global groundnut production is in the semi-arid tropics, ICRISAT has identified plant types and traits with increased drought tolerance. Improved varieties are expected for release to farmers in 2000.


Pigeonpea is important in Southern Asia and Eastern Africa. In addition to water deficits and water logging, yields are affected by a variety of pests and diseases which cause some S1.1 billion in avoidable losses per year.

Wilt disease alone accounts for $193 million annual loss, according to ICRISAT. After identifying resistance in germplasm, ICRISAT bred a range of genotypes for most cropping systems that include pigeonpea at risk of this fungal pathogen. Four varieties have been released and are already grown on one-fifth of Malawi's pigeonpea area. Wilt disease-resistant varieties are expected to spread quickly.

Discovery of cytoplasmic male-sterility allowed ICRISAT to develop pigeonpea hybrids with a 40 percent yield advantage over conventional varieties, permitting the production of cheaper hybrid seed. By 2002, millions of poor are expected to benefit in terms of better nutrition and health.


Farmer cowpea yields are still very low, at less than one-tenth of research station yields. Cowpea is best suited for the semi-arid tropics and moist savanna. IITA has developed some cowpea varieties resistant to striga, a parasitic weed, that are now being grown. Ongoing research in cowpea biotechnology should produce varieties resistant to various pests.


Soybean is a recent introduction to Africa. IITA has made excellent progress in developing soybean products that fit into African diets and food habits. Soybean is best adapted to moist savanna zones but does not grow well on acid soils. Frogeye leaf spot and red leaf blotch are the major biotic constraints.

With the expansion of production, increased insect problems are being expected, and IITA is working on the timely development of appropriate solutions.


Lentil is a crop adapted to low rainfall, dry areas in the Middle East and North Africa and the rice-based farming systems in Asia and Latin America. Besides providing high-quality protein, it also produces high-quality straw, which forms an integral part of the feed supply to small ruminants in MENA. ICARDA, which has a mandate to improve this crop, has developed high-yielding varieties with greater total biomass yield, drought tolerance and resistance to disease. The incorporation of resistance to pod shattering, lodging resistance and greater plant height has increased the yield potential of the crop. Mechanical harvest of such varieties can reduce production cost and encourage farmers to expand production. Large production increases have occurred in Turkey and Asia through collaborative research with national programs. Sudan has recently become self-sufficient in lentils.


Faba bean is a staple in the diets of people in the Middle East and North Africa, China and several Latin American countries. After chickpea and pea, this is the most important cool-season food legume worldwide. The low-income people in the Nile Valley countries and North Africa use faba beans at least once a day in their cereal rich diet. The yield and yield stability are constrained by foliar diseases and parasitic weeds. ICARDA has identified with national programs sources of resistance to this problem and incorporated these resistance traits in well-adapted varieties. Such varieties have been released and now cover large areas in Egypt, Sudan, Ethiopia and China. Employment of these varieties has resulted in a considerable reduction in the use of pesticides in these countries.

Return to Table of Contents