Development and spread of improved maize varieties and hybrids in developing countries

Material Information

Development and spread of improved maize varieties and hybrids in developing countries
Timothy, David H ( David Harry )
Harvey, Paul Henry, 1911-
Dowswell, Christopher R.
United States -- Agency for International Development. -- Bureau for Science and Technology
Place of Publication:
Washington DC
Bureau for Science and Technology, Agency for International Development
Publication Date:
Physical Description:
xi, 71 p. : ill. ; 28 cm.


Subjects / Keywords:
Corn -- Breeding -- International cooperation -- Developing countries ( lcsh )
Hybrid corn -- International cooperation -- Developing countries ( lcsh )
bibliography ( marcgt )
federal government publication ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references.
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
Statement of Responsibility:
David H. Timothy, Paul H. Harvey, and Christopher R. Dowswell.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.
Resource Identifier:
19287757 ( OCLC )

Full Text
Development and Spread of
Improved Maize Varieties and Hybrids in Developing Countries
for Ater ational Development
ik f
*1=_ N.
' t /
..:Irl N

Maize is the third most important cereal crop in the developing world, after rice and wheat. Attempts have been made over many years to improve maize varieties and hybrids used in developing countries. National and international programs have been involved. In the latter case, research has been concentrated at the International Maize and Wheat Improvement Center (CIMMYT) in Mexico and the International Institute of Tropical Agriculture (IITA) in Nigeria. Both centers carry out their work in collaboration with national research programs.
Although the adoption of high-yielding wheat varieties developed by CIMMYT and of rice varieties developed by other centers has been well documented, the same is not true of maize. Part of the reason is that the genetics of maize are quite different from, and much more complicated than, wheat or rice. Improved varieties are not as easily identified visually. Also, relatively little effort has been put into the measurement process. As a result, it is not well known how much impact the maize research programs have had at the farm level.
This report is a first step in rectifying this information gap. It attempts to identify the maize varieties, hybrids, population, pools, and lines developed by CIMMYT and IITA that are being used in national programs in developing nations. It also incorporates, where available, estimates of area planted with hybrids and improved varieties. Further work is needed to more fully document the main gene pools and to verify and expand the statistical data at the national level.
Key Words
Maize, corn, maize/corn breeding, maize/corn varieties and hybrids, agricultural research, international agricultural research centers, developing countries
Cover photograph from CIMMYT

David H. Timothy, Paul H. Harvey, and
Christopher R. Dowswell
Bureau for Science and Technology
Agency for International Development
Washington, DC
'I~jjiji 1988

David H. Timothy is Professor of Crop Science, Botany, and Genetics at North Carolina State University, Raleigh. He served as a plant geneticist for the Rockefeller Foundation in Bogota, Colombia, from 1956 to 1961, where he headed the Cooperative Maize Improvement Program of the Foundation and the Ministry of Agriculture. He also worked closely with maize improvement programs mn other Latin American countries. During 1985-86, Dr. Timothy was Chief Scientist of the U.S. Department of Agriculture's Competitive Research Grant Program in Washington, D.C.
Paul H. Harvey was William Neal Reynolds Professor and head of the Crop Science Department at North Carolina State University, Raleigh, until his retirement in 1975. He then served as coordinator of corn, grain sorghum, and grain millet for the National Program Staff of the Agricultural Research Service (ARS) of the U.S. Department of Agriculture in Beltsville, Maryland, until 1980. In the early 1980s, Dr. Harvey served as coordinator for the National Corn Research Needs and Priorities Study. In addition to his domestic work in corn breeding, he has served as a consultant overseas.
Christopher R. Dowswell is a consultant in agricultural communications residing in Mexico. He is currently coauthoring a book titled Maize in the Developing World: A Handbook for Research Administrators. He was formerly head of Information Services at the International Maize and Wheat Improvement Center (CIMMYT) from 1978 to 1985, and at the International Fertilizer Development Center (IFDC) from 1976 to 1978. He also worked as an extension agent and communications specialist for Oregon State University during 1973 to 1976.
The views expressed in this bulletin are those of the authors and are not necessarily those of the U.S. Agency for International Development. Mention of commercial firms and/or their products does not constitute or imply endorsement.
Publication development services provided by
Washington, D.C.

Foreword v
Preface vii
Summary ix
1. Background on Maize 1
Maize in the World Economy 1
Area and production I Utilization 1
Some Basic Biological Characteristics of Maize 3
Botanical classification 3 Climatic adaptation 3
Influences on Maize Improvement 3
Selection criteria and plant structure 3 Selection criteria and use 3 Selection
criteria and population structure 4 Infrastructure and plant breeding 4
2. International Germplasm Development Programs and Networks 7
Early Maize Germplasm Efforts and Approaches 7
Germplasm collections in Latin America 7 Development of improved varieties
and hybrids in Latin America 8 Maize research activities in Asia 9
Exchanges of breeding material and informal cooperative networks 10
Emerging concepts of germplasm use and breeding methods 10
Refinements of population improvement methods 11
Establishment of International Maize Improvement Programs 11
CIMMYT's international maize improvement program 13
IITA's maize improvement program 15
3. Recent Releases and Use of Improved Maize Materials 19 Sub-Saharan Africa 19
Eastern and Southern Africa 19
Ethiopia 19 Kenya 19 Lesotho 21 Malawi 21 Mozambique 23
Reunion 23 South Africa 23 Somalia 24 Swaziland 24 Tanzania 24
Uganda 26 Zambia 26 Zimbabwe 26
Western and Central Africa 28
Benin 28 Burkina Faso 28 Cameroon 28 Central African Republic 28
C6te d'Ivoire 28 Gambia 29 Ghana 29 Liberia 29 Nigeria 30
Senegal 30 Sierra Leone 31 Togo 31 Zaire 31

North Africa and the Middle East 32
North Africa 32
Egypt 32 Morocco 32
Middle East 33
Turkey 33
Asia 34
South Asia 34
Burma 34 India 34 Nepal 35 Pakistan 36 Sri Lanka 37
Southeast Asia and Pacific 38
Indonesia 38 Philippines 38 Thailand 39 Vietnam 41
East Asia 41
China 41
Latin America 42
Mexico, Central America, and Caribbean 42
Costa Rica 42 Cuba 42 El Salvador 42 Guatemala 43 Haiti 44
Honduras 44 Mexico 44 Nicaragua 46 Panama 46
Andean Region, South America 46
Bolivia 46 Colombia 46 Ecuador 47 Peru 48 Venezuela 49
Southern Cone, South America 49
Argentina 49 Brazil 49 Chile 50 Paraguay 51
4. Summary of Use of Improved Genotypes 53
CIMMYT Survey of Improved Maize Seed Use 53
Regional Totals 54
Eastern and Southern Africa 55 West and Central Africa 57 North Africa a
the Middle East 57 South Asia 58 Southeast Asia and the Pacific 58
East Asia 59 Mexico, Central America, and the Caribbean 59 Andean
countries of South America 60 Southern Cone countries of South America 61
Accelerating the Use of Improved Maize Genotypes in the Third World 61
References 63
Published References 63
Correspondence and Interviews 66
Manuscript Reviews 67
Appendix A. A Note on Some Maize Breeding Terms 69
Appendix B. Yield Potential of Tropical Hybrids in the United States 71

This report had its origins in late 1983. At that time, the Consultative Group on International Agricultural Research (CGIAR) organized an Impact Study of the research it had sponsored in developing nations. As part of the commodity portion of that effort, there was interest in initiating a study on the development and adoption of improved maize germplasm.
There was good reason for this interest. Maize is a major crop in developing countries, and it represents a major line of activity at two centers sponsored by the CGIAR-the International Maize and Wheat Improvement Center (CIMMYT) in Mexico and the International Institute for Tropical Agriculture (IITA) in Nigeria. While research on maize has been underway at both centers since their inception, relatively little has been done to trace the adoption and use of maize technology produced by these centers in cooperation with national programs.
The Agency for International Development was requested to provide some in-kind assistance to the CGIAR Impact Study. I was asked to update my earlier studies on the development and spread of high-yielding varieties of wheat and rice (Dalrymple 1986a,b) and some thought was given to having me initiate a similar effort on maize. The latter step, it quickly became apparent, was not possible-both because of time constraints and the need for technical knowledge of maize genetics and breeding. Hence, it was decided to contract the maize portion to a maize specialist. Dr. Nyle C. Brady, A.I.D.'s Senior Assistant Administrator for Science and Technology, was keenly interested in initiating the work and saw to it that the necessary financial resources were available for the study. The Office of International Cooperation and Development of the U.S. Department of Agriculture assisted in the search and engagement of suitable individuals to carry out the work.
Arrangements were made with Drs. David Timothy and Paul Harvey of North Carolina State University in the spring of 1984 to initiate the work. They produced a first draft by early 1986 (Dr. Timothy was primarily responsible for chapters 1 and 2 and Dr. Harvey for chapter 3). The manuscript was then distributed for technical review. In the course of this process, a considerable amount of new country information became available, principally from CIMMYT, but also from IITA and the CGIAR Impact Study itself. Unfortunately, by then, the authors had used up the time they had available for the project.
Christopher Dowswell, formerly head of the CIMMYT information office and now a consultant in agricultural communications, was subsequently engaged to incorporate this new information. He started work in early 1987 and acquired additional information from CIMMYT and IITA personnel and from national maize researchers in several Asian and African countries in the course of travels to those nations on other assignments. He also interacted with Drs. Timothy and Harvey. The revised manuscript was then reviewed again and further modifications were made; this process was completed by June 1988.
Thus, the final report is the product of a two-stage process stretching over 4 years. I think that the two stages-involving individuals with different backgrounds-introduced some "hybrid vigor" into the report. This vigor has been needed because of the pioneering and difficult nature of the enterprise. We now better understand why it has not been done before. At the same time, we recognize the importance of the task and the need to record some of the information before it disappears.

There is, inevitably, some unevenness in the country coverage in the report. Principal emphasis is placed on the distribution and use of germplasm provided by CIMMYT and IITA. Obviously, the related contributions of national programs would be given further attention in a more comprehensive treatment. More is also said about public rather than private research efforts. This is because most maize research in developing countries has been conducted by public-sector organizations and because less information is available from private-sector organizations on the genetic resources used in genotype development. Areas where the private sector is particularly active, such as the Southern Cone of South America, are not covered to the extent that their importance in maize production might have suggested. This is particularly true of Argentina, for which very little information is reported. Africa, on the other hand, receives extensive coverage-reflecting the extent to which maize improvement research is a public sector effort and, in part, the large number of countries with maize research programs.
The country presentations are, in any case, quite brief and tend to focus on the development and commercial release of improved varieties and hybrids rather than the actual coverage and impact on yields of these improved genotypes at the farm level. In most cases, it has not been possible to provide much statistical information, particularly of a time-series nature, about the area planted with improved maize varieties and hybrids in individual countries.
The uninitiated might tend to think of maize breeding as a passionless, scientific enterprise. It certainly is scientific, but there are-as when one scratches the surface in other areas-strong differences in opn**on among some scientists. The study attempted to strike a middle ground and stay with the facts as closely as possible.
There is plenty of work left to do-both to expand the breadth and depth of the coverage. I hope that this report serves both as an introduction to the subject and as a stimulus for further study.
Dana G. Dalrymple
Project Advisor
Directorate for Food and Agriculture
Bureau for Science and Technology

This first report on the development and use of improved maize genotypes in the Third World is modeled after the reports prepared periodically by Dana G. Dalrymple on the development and spread of high-yielding varieties of wheat and rice in the less developed nations (Dalrymple 1986ab). While we are disappointed that this report is not as complete in scope or data coverage as those reports, we are pleased in the belief that it will be a useful resource for subsequent efforts.
Our original hope was to obtain reasonably accurate estimates of the use of improved, specifically identified germplasm, varieties, or hybrids emanating from the International Maize and Wheat Improvement Center, headquartered in Mexico and commonly known by its Spanish acronym, CIMMYT, and the International Institute of Tropical Agriculture (IITA), headquartered in Nigeria.
We planned to establish the genealogies of the varieties, hybrids, populations, pools, and lines being developed by these two international agricultural research centers. Having the known genealogies in hand would permit (1) more reliable estimates of the germplasm contribution of these two IARCs to national program efforts to increase maize production in the Third World, (2) more definitive concepts of which basic germplasm is most noteworthy in that production, and, perhaps, (3) identification of basic germplasm that might receive more attention in future maize improvement strategies.
The peculiar (as opposed to wheat or rice) characteristics of maize biology and its culture have complicated efforts to provide an accurate picture of the use of improved varieties and hybrids in developing countries. Nonetheless, we have assembled a sizable amount of anecdotal information, oral histories, copies of lists and unpublished reports, and copies of genealogies and pedigrees made from field books about breeding material extracted from, or combined with, other germplasm. But informative and fascinating as these particulars appear, their range in accuracy, completeness, precision, and scope will require additional investigation. Subsequent efforts might well concentrate on ferreting out more reliable information on cultivars or material being developed having known genetic relationships to germplasm from the LARCs.
There is a particularly strong need to investigate the origins and breeding methodologies used to develop the outstanding germplasm complexes that have contributed so much to the improvement of maize. Many of the detailed records on how these germplasm complexes were put together and what mating and selection systems were employed are still available from field books and unpublished reports. Some of the details are missing, however, and attainable only from those researchers and breeders who put the materials together in the first place. Many of these people have retired and some have died. Thus, it is unlikely that the necessary records about the origins and handling of these germplasm complexes will be preserved much longer. If the formation of these outstanding gene pools and populations and the methodologies used to improve them are to be documented, it must be done promptly.
The original draft of this study has been considerably modified through the additional information and comments provided by CIMMYT and IITA and by several national maize programs in Asia, the Middle East, and Africa. Several CIMMYT and IITA maize publications released in 1986 and 1987 have provided new information on germplasm development

in developing countries. National production statistics published by FAO have been added in chapters 1 and 4. The statistics on the use of improved genotypes come from a CIMMYT survey conducted in 1985-86 on commercial seed production in developing countries (CIMMYT 1987). Even with these data additions, there is still inadequate information on germplasm development and use for many countries.
Numerous maize researchers in developing and developed countries have given generously of their time in interviews and correspondence. We are most grateful to them. Various drafts of this report were reviewed by research leaders at CIMMYT, IITA, and several private plant breeding companies. To them we extend our profound thanks. A special expression of gratitude goes to Dana G. Dalrymple for his encouragement, guidance, and thoughts throughout this project.
David H. Timothy
Paul H. Harvey
Christopher R. Dowswell

Maize is the second most important cereal crop in the world overall, after wheat and before rice. In the developing world, maize ranks third, after rice and wheat. In total, there are 80 million ha. planted with maize in developing countries. This represents 60% of the world's maize area, though only 40% of global production is harvested from Third World maize lands. Tropical and subtropical environments cover 65% to 70% of this area and temperate environments cover the remainder. Four countries account for 67% of total Third World maize production: China, Brazil, Mexico, and Argentina.
The biology of maize and its population structure are vastly distinct from those of wheat and rice. The ramifications of these differences are reflected in (1) the use of breeding materials and methodologies, (2) technology and capability of seed production and distribution, and (3) local customs and conditions pertinent to grain production and consumer uses of maize. Improvement of maize germplasm for local, national, or international programs may be accomplished by emphasis on population improvement, concentration on hybrids and inbred lines, or modifications of either or both approaches.
Maize research in developing countries received an important impetus from the collaborative maize improvement programs established during the 1940s and 1950s by the Rockefeller Foundation and the governments of various developing countries. One of the first maize research efforts supported by the Rockefeller Foundation was the collection, classification, and preservation of the native maize races in Latin America and the Caribbean-centers of origin for maize. The best germplasm collections were assembled into breeding populations and improved for yield and various agronomic traits; these populations were then shared with selected national programs in the Americas, Asia, and Africa. A handful of these germplasm complexes introduced substantial new amounts of useful genetic diversity into national maize research programs, especially in Asia and Africa, where the genetic base of national breeding materials had become very narrow. These germplasm introductions accelerated progress in many maize breeding programs, and literally hundreds of improved varieties and hybrids have been developed using germplasm obtained through international exchange.
The International Maize and Wheat Improvement Center (CIMMYT) and the International Institute of Tropical Agriculture (IITA), established in the mid-1960s, represent the "institutionalization" of many of the international research and training activities pioneered during the 1940s and 1950s by the Rockefeller and Ford Foundations, FAG, and USAID. During the past two decades, the international agricultural research center (LARC) system has greatly expanded the amount and scope of international agricultural research. Similarly, developing countries, impressed by the green revolution gains in wheat and rice, have markedly expanded their investments in national agricultural research systems, including maize improvement research.
Today, CIMMYT and IITA serve as research hubs of large maize improvement networks involving hundreds of maize scientists in over 100 developing countries. Both IARCs operate large (and growing) maize research programs. Through international testing programs, vast amounts of improved maize germplasm are distributed for evaluation and use by national breeding programs. Each year, tons of seed for experimentation are shipped by CIMMYT

and IITA to hundreds of national collaborators throughout the world. Both IARCs make their seed freely available to bona fide maize researchers; preference, however, is given to collaboration with public sector maize research organizations. Of the two IARCs, CIMMYT has the most widespread maize improvement program, with collaborative relationships with more than 80 countries in Latin America, Asia, and Africa. IITA has maintained a regional focus in its maize improvement work, concentrating on the germplasm requirements of sub-Saharan Africa.
The complex genetic backgrounds of the new maize types developed, or being developed, preclude a simplistic compilation of data that would attribute increased yield or production to the use of CIMMYT and IITA materials. The improved maize germplasms developed and/or distributed by CIMMYT and IITA do not have a single common characteristic-such as the dwarfing habit so obvious in rice and wheat-to facilitate the tracking of production data. Even so, this report shows that developing countries have released hundreds of varieties and hybrids based on CIMMYT and IITA germplasm, with CIMMYT being a more dominant germplasm contributor. IITA's streak-resistant germplasm has been extremely valuable to national maize programs, especially in sub-Saharan Africa. Some of IITA's tropical inbred lines and hybrids are also showing promise in maize production zones in sub-Saharan Africa and on other continents. In some instances, increased production from the use of CIMMYT and IITA maize stocks has been noteworthy. In other cases, the results are nebulous at best.
In the temperate areas and, to some extent, in the subtropical areas of the developing world, germplasm from the USA and Europe has been widely used to develop improved genotypes. In the subtropical areas, however, temperate germplasn has generally lacked sufficient resistance to stalk and ear rots and various foliar diseases. To overcome this deficiency, tropical and subtropical germplasms with greater resistance to these diseases have been introgressed to strengthen the disease resistance of genotypes based on temperate germplasm.
Improved genotypes have been developed and released for most lowland tropical and subtropical areas of the developing world. These materials have higher yield potential and superior agronomic characteristics than traditional maize varieties. The improved genotypes are shorter in stature than traditional local materials and partition more of their total dry matter production to grain. Their resistance to foliar diseases and stalk and ear rots is also generally superior to traditional varieties. There is still a need to develop genotypes for intermediate and highland environments and for those areas characterized by serious environmental stresses, especially drought.
CIMMYT has estimated that in 1985-86 approximately 50% of 80 million ha in developing countries was planted with commercial seed of improved hybrids and improved varieties. In China, Brazil, and Argentina-which have 30 million ha of maize land-commercial seed is used on over 75% of the total maize areas. However, when these three countries are not included in the statistics, only about 26% of the total maize area in developing countries is planted with commercial varieties and hybrids. While improved genotypes have been developed that are suitable for much of the area still growing local varieties and landraces, other impediments-especially nonavailability of commercial seed and fertilizer-restrict their adoption.
Although improved genotypes have been developed and released for most of the major production environments, the absence of effective seed production systems has seriously restricted the diffusion of these materials at the farm level. Unfortunately, in all too many countries, the absence of functioning maize seed sectors has severely restricted the diffusion of improved varieties and hybrids at the farm level. This, in turn, has seriously reduced the returns that would have otherwise accrued to society from the investments made by developing country governments in maize improvement research. Clearly, nonfunctioning maize

seed sectors in all too many developing countries have had a tremendous cost in terms of foregone maize production and improvements in productivity.
Since the late 1970s, private sector involvement in maize research and seed production in the developing world has grown markedly. Initially, private maize seed companies concentrated their activities in temperate-zone developing countries where U.S. Corn Belt hybrids do well. Increasingly, however, as more improved germplasm for tropical and subtropical areas has become available (developed largely through national and international public sector research efforts), private sector maize research and seed production initiatives have expanded in those countries with subtropical and tropical maize production environments where strong commercial demand exists for maize.

MAIZE IN THE WORLD area. Highland zones above 1,500 m (subtropECONOMY ical-temperate) in Mexico, Central America, Andean
countries, East Africa, and Himalayan countries account
Area and Production for about 6 million ha (8%) of the total maize area in
the Third World.
Maize, with a global harvest in 1985 of 449 million During the period 1961-65 to 1983-85, world maize metric tons from 133 million ha, ranked second to production increased by 182 million t (CIMMYT 1987). wheat (with hulled rice third) among the world's cereal This is a 4.0% annual growth rate and represents a 84% crops (FAO 1986). Some 70 countries produce maize increase in world supplies. Most of this growth was on 100,000 ha or more; 53 of these are developing achieved through yield improvements. In developing countries. Developed market economies account for countries-though there was considerable regional 30% of the global maize area but provide 50% of total vaito-elsncaedta2.%nulrtwhe
prouctondu toaveag yildstha ae treet*es the overall area expanded by 1.2% per year. China higher than the world average. Developing countries rprted the mostrpdgot nyed mn ao
accont or aproimatly 0% o th totl wrld producers, with a 4.8% per annum rate of gain in yield maize area but produce only 40% of the global harvest, per year. In sharp contrast, yields in sub-S aharan During 1983-85, developing countries produced an Africa increased at only a 0.9% yearly rate. For the average of 169 million tons (t)l of maize per year developing countries as a whole, production is not (CIMMYT 1987). Four countries accounted for 67% increasing as fast as demand, and imports are rising of the Third World's maize production: China (68 (Vocke 1987). million t), Brazil (21 million t), Mexico (14 million t),
Argentina (10 million t). Utilization
In total, there are 80 million ha planted with maize
in developing countries. Lowland tropical environments Maize is used in more ways than any other cereal: account for 32 million ha or about 40% of the total area as a human food, a feed grain, a fodder crop, and for planted with maize in the developing world. These hundreds of industrial purposes. Its grain, stalk, leaves, tropical maize growing environments are found in the cobs, tassel, and silks all have commercial value in most lowlands of eastern and western Africa, south of the settings, though that of the grain is the greatest. The Sahara; the plains and delta areas of South Asia; most diversified use of maize occurs in the United Southeast Asia; Central America and parts of Mexico; States, where over 1,000 products in a typical superand the lowland areas in Andean countries of South market contain maize in some form or another in their America and Brazil. Temperate environments (e.g., makeup. Argentina, Chile, China, Middle East), accounting for Worldwide, about 66% of all maize is used for 24 million ha (30%) of the total maize area, are in feeding livestock, 25% for human consumption, and developing countries. Subtropical and intermediate- 9% for industrial purposes and as seed (CIMMYT elevation zones (South Asia, Middle East, North Africa, 1984). In the developing world, however, roughly 50% Mexico, Brazil, East Africa) account for approximately of all maize is consumed by humans as a direct food 18 million ha (22%) of the developing country maize source, 43% is for livestock feed, and the remainder is
______________used for industrial and seed purposes (CIMfMYr 1984).
1Al1 ton measurements are for metric tons. Although maize is important as a food crop in Mexico,

Figure 1. Cultivating maize interplanted with other crops in Honduras. Source: CIMMYT.

CentralAmerica and the Caribbean, the Andean coun- imposed selection pressures on the kinds of maize at tries of South America, and sub-Saharan Africa, it is each site. increasingly important as a feed grain in the newly
industrialized countries of the Pacific Rim, the INFLUENCES ON MAIZE middle-income countries of Latin America, and the IMPROVEMENT major oil-exporting countries. In the industrialized societies, maize is bred for
SOME BASIC BIOLOGICAL mechanized agricultural technologies, highly engineered
CHARACTERISTICS OF MAIZE milling and processing plants, sophisticated marketing
strategies, and various consumer tastes and choices.
These industrial endeavors are supported by, and are
Botanical Classification part of, a well-functioning and highly developed infrastructure. However, in those areas in which sufficient
Maize, Zea mays L. spp. mays, is a diploid (2n= 20) levels of development do not exist, the farmer has member of the tribe Maydeae, Tripaceae, or Andropo- played an important role in the evolution of the kinds neae (depending upon one's taxonomy), subfamily of maize being grown. This is particularly so when Panicoideae of the Gramineae. Its native distribution maize is used for direct human consumption rather was restricted to the New World. Although the origins than as animal feed or industrial use. of maize remain elusive, it is well established that at
least one kind of maize originated in the Valley of Selection Criteria and Plant Structure Mexico (Galinat 1971, 1977; Mangelsdorf et al. 1964, The structure of maize lends itself to stringent and 1967; Mangelsdorf 1974). The archeobotanical evi- repeated scrutiny of each individual plant by its cultivadence indicates that for several thousand years the tors. Thus, selection of a desired type may be of pencil thin cobs of this primitive plant nourished its exceedingly high intensity. This kind of selection is cultivators. Subsequently, new characteristics were added facilitated by the large physical structure of the plant; to this primitive cultivar, presumably from the related all the above-ground parts are easily observed for wild grass teosinte, and the domesticated genotypes length, width, thickness, color, pubescence, position, began to assume the proportions and characteristics of and numbers of plant parts and their shapes, angularity present day maize land races. While reasonably close, or idealized architecture, husk cover, shank characterthe affinity of maize with Tripsacum L. is much less istics, maturation rates, and reaction to insects, disthan with the teosinte. However, there is little doubt eases, or other stresses and perturbations. that maize, teosinte, and Trisacum share differing The ear is particularly susceptible to selection crikinds and amounts of common genetic information, teria. Each ear is harvested by hand and through repeated handling in the husking, drying, storing and
Climatic Adaptation shelling process, human preference comes to bear for
grain type. These traits include grain size, shape, color,
Maize is grown in more diverse areas of the world starch texture, characters pertaining to appearance, than any other major crop. It is grown from sea level to cleanliness, and desirability for ultimate use. The suc3,800 m elevation near Lake Titicaca in Bolivia and cesses of the efforts are evident from the extant variPeru, from desert oases to zones having 11,000 mm ability among the thousands of indigenous strains of rainfall (Patifio 1956) along the western coast of Co- maize found throughout the world. lombia, and from about 420 latitude S near Chiloe
Island to about 50* latitude N on the Gaspe Peninsula Selection Criteria and Use
of New Brunswick, Canada. It is cultivated from north- Superimposed upon the selection pressure of enviern Europe and Russia to South Africa, eastward
through Asia, the Himalayas, China, Southeast Asia, ronents or biological stresses, such as temperature or and the Pacific Islands. The genetic differences are vast insects, are the additional selection criteria pertinent to among the kinds of maize grown among the distinct the local uses of the grain or other plant parts. Maize locales of these disparate areas. The diverse but par- is used as a staple in many forms of comestibles: ticular conditions of soil, temperature, rainfall, relative fermented and nonfermented drink, porridges, tortilhumidity, photoperiod, and light intensity have all las and other baked products, cooking oil, dyes, animal

Figure 2. Preparing tortillas from maize in a Latin American home. Source: CIMMYT.
and poultry feedstuffs, fodders, medicinals, ceremo- brid-occurs and the approach to homozygosity of the nial and religious essentials, physical support of a progeny from any heterogenous individual (the F1) companion crop in multicropping systems, toys, dolls, increases by one-half in each successive generation. effigies and other household or community symbol- However, the population will then contain a mixture of isms, and raw material for a host of industrial products. distinct homozygous lines and a decreasing number of heterozygous individuals.
Selection Criteria and The monoecious character of maize (separate male
Population Structure and female inflorescences on the same plant) facilitates controlled pollination. Thus, many of the breedMaize is a highly evolved cross-fertilizing annual. ing methods used for self-pollinating species are used Generally, each plant is an indigenous or intermating with maize for certain qualitative traits and conditions, entity that is unique by virtue of its hybrid nature, such as simply inherited reactions to diseases. Quantihaving originated from a male gamete of one plant and ticharacteried an tdkidssofdomiant, a female gamete from another. While the plants in such native characters (yield) and the kinds of dominant, a population maybe phenotypically homogeneous as a additive, or interactive gene action controlling such consequence of selection, the individual genetic condi- traits and hybrid vigor have led to the development of numerous breeding methodologies. These have been
tion of each plant is essentially heterozygous. How- used to develop inbred lines for hybrids, synthetics, ever, when two completely homozygous inbred lines copstorirvevaeisndfroulin are used to form an F1 hybrid, all the progeny in that composites, or improved varieties, and for population first generation will be genetically identical. Subse- improvement as source materials. (See appendix A for quent generations will revert to individually distinct a brief definition of hybrids, synthetics, and composheterozygous plants comprising all the possible combi- items ) nations of the alleles of genes from the inbred lines.
This is distinct from what occurs in self-pollinating
species, such as wheat or rice, in which a population is Sustained production of high-yielding hybrids largely homozygous: each plant is identical to another. requires an adequate technological and agricultural In some self-pollinations, an outcross-an F1 hy- infrastructure comprised of credit, transportation, proc4

essing, marketing, educational, and other agricultural fled hybrids, synthetics, or varietal crosses. These are support systems. Areas lacking these facets are either usually produced by governmental or international unsuitable or less developed for hybrid production. In agencies mandated by agricultural development needs these latter cases, maize is usually produced from rather than the requirement to make a profit. These indigenous races, improved varieties, synthetics, or latter agencies also develop single crosses or other from varietal crosses. Indeed, most national programs hybrids for commercial production, but when seed in the developing world produce improved maize sales volume reaches an appropriate level, commercial genotypes via these latter systems rather than by the companies soon appear and target hybrid seed sales on use of single, three-way, or double-cross hybrids, more specific areas of their product's adaptability than
Commercial breeding efforts are sustained only in is normally done by a development agency.
the more advanced maize growing areas, as investment The presence (or absence) of effective seed proreturns are made only by repeated seed sales of hybrid duction organizations is the major determinant of the seed. Where farmers save or sell seed from a hybrid for type of maize varieties used by farmers around the replanting, yield potential decreases appreciably, and world. Where adequately developed seed industries agronomically undesirable recombinant characteris- exist, F, hybrids are generally the "standard." Where tics are expressed in the next segregating generation. seed industries are less well-developed and/or where Neither the farmer nor the marketplace will tolerate farmers grow maize as a subsistence crop for home this. Thus, maize improvement programs in less ad- consumption, open-pollinated varieties, often unimvanced areas are based on improved varieties, modi- proved, are the dominant types.

EARLY MAIZE GERMPLASM lar procedures were subsequently used in other areas
EFFORTS AND APPROACHES to describe on a preliminary basis the races of maize in
the NewWorld (Brieger et al. 1958; Brown 1960; Grant
The rapid spread of hybrid maize production in the et al. 1963; Grobman et al. 1961; Hatheway 1957; United States during the 1940s was accompanied by Ramirez et al. 1960; Roberts et al. 1957; Stakman et al. the disappearance of indigenous maize varieties. The 1967; Timothy et al. 1961, 1963; Wellhausen et al. loss of those genetic resources, products of hundreds 1957). or thousands of years of evolution under domestication, could restrict future maize improvement and b Over 280 races were described, with some overlap, deprive us of genes of great potential importance. but for the first time the variation in an important world There was a growing realization that the genetic diver- crop was classified into easily recognized and workable sity among native strains and varieties in Latin Amer- groups. Additional races in both the New and Old ica and the Caribbean would also become extinct with Worlds have been described subsequently, but these increased planting of improved varieties and hybrids studies are generally of restricted geographical scope. that had begun in those areas. Concern for this situ- (See Brown and Goodman 1977; Goodman 1978; ation led to concerted efforts beginning in the 1940s Hallauer and Miranda Fo 1981.) and 1950s to collect, preserve, and classify native maize The similarity of many of the races and their evoluvarieties and strains in Latin America. These early tionary histories have led to additional subsequent efforts in maize germplasm collection and classifica- groupings of races into groups or racial complexes, tion led to the identification and development of some mostlyby Goodman andhis coworkers. (See Goodman outstanding germplasm complexes in the Americas, 1978 and Brown and Goodman 1977.) The identificawhich later also proved to be veryvaluable to the maize tion of these complexes, coupled with the results and breeding programs in Asia and Africa. experience from plant breeding studies, has begun to
enable the identification of superior kinds of germplasm
Germplasm Collections in for rapid improvement of yield and other characterisLatin America tics.
The Cooperative Maize Improvement Program of The vast array of diversity in this material, the
the Mexican Ministry of Agriculture and the Rockefeller specific requirements for adaptation to local environFoundation, initiated in 1943, immediately began col- ments, demands by local markets, and preferences lecting the native maize varieties and strains through- dictated by human consumption added impetus to the out Mexico to serve as a germplasm base for improve- need for collections. The obvious similarity of certain meant. The myriad kinds of variability in this hodge- types of maize in the Caribbean Basin to those of the podge of diversity was reduced by several criteria into eastern coast of Mexico, and of maize in northern and small groups containing similar or like kinds of recog- western South America to either western Mexico or nizable traits. These smaller groups were defined and southern Mexico and highland Central America, indicharacterized as races (Wellhausen et al. 1952). Simi- cated that collections outside Mexico would also be

Figure 3. Genetic diversity of maize is expressed in these highly varied ears. Source: CIMMYT.
essential. Similarly, the Cooperative Maize Improve- ment. The Central American and Caribbean collecment Program of the Colombian Ministry of Agricul- tions were to be maintained and distributed at Chapture and the Rockefeller Foundation begun in 1950 ingo, Mexico; those from western and Andean South reiterated the need to collect maize beyond the Colom- America from Venezuela through Chile at Medellin, bian border. Activities by commercial seed companies Colombia; and those from eastern South America and and universities in the industrialized countries during the Guianas at Piracicaba, Brazil. As a safeguard this time period also stressed the need for collection against loss, duplicate samples of the collections were (Brown 1960; Cutler 1946; Hatheway 1957). to be sent through the USDA facilities at Glen Dale,
The National Academy of Sciences-National Re- Maryland, and eventually put in long-term storage at search Council (U.S.) formed the Committee on Pres- the National Seed Storage Laboratory at Fort Collins, ervation of Indigenous Strains of Maize to sponsor a Colorado. However, the reality of those events has project that would collect, preserve, and study for been less than satisfactory (Goodman 1984; Timothy future use as many land races and varieties of the and Goodman 1979). Americas as possible (Clark 1954; 1956). Additional Development of Improved Varieties support and cooperation for this work was received and Hybrids in Latin America
from the Rockefeller Foundation, numerous governments, and several universities. The collections were During the 1940s and 1950s, several national maize made, studied, and described in a "Races of Maize" improvement programs made appreciable impact in bulletin series. These germplasm collections were to the development and distribution of improved varieties be maintained and made available to qualified scien- and hybrids. The Rockefeller Foundation initiated a tists for additional study, research, and maize improve- Central American Corn Improvement Project in 1954

(Stakman 1967). Particularly notable national maize lected for immediate distribution and as foundation breeding programs were found in Mexico, Colombia, material for later development of superior synthetics Cuba, Venezuela, Brazil, and Argentina. Materials and hybrids. The improved yield of these materials from these programs made considerable contributions ranged from 15% to 160% above comparable local to breeding stocks and are still being used. varieties (Wellhausen 1947).
An improved yellow flint variety, Venezuela-i, was The Mexican-Rockefeller research group develdeveloped by several cycles of mass selection from two oped the composite variety, Rocamex V-7, by mass Cuban varieties (Langham 1942). Although not stated, selections in a collection from near Actopan, Hidalgo, this Coastal Tropical Flint has Cuban Flint Cateto in its originally arising from a natural cross of cylindrical background. The development in Colombia of several dents from the lowland tropics with high-altitude maize varieties, Colombia 1, Colombia 2, and ETO (Chavar- (conicos) from the Central Plateau (Wellhausen and riaga 1966), were of considerable worldwide impor- Roberts 1948). Several other varieties were developed tance. That germplasm is an important component of ts 14)eveRoter eties were deloped
many breeding programs of the world. using this procedure (Roberts et al. 1949; Wellhausen
many b ian olo m bia of t e l. andwh 1950), as were those based on improved collections of
Colombia 1 and Colombia 2, yellow and white the Race, Olot6n (Neves et al. 1957). varieties, respectively, were both developed from a
series of crosses among three sources followed by Early hybrids released bythe Mexican-Rockefeller selection for combining ability among sib-mating/ Program for tropical regions were Rocamex H-501, progenies. The three sources of germplasm were the H-502, H-503 (Reyes et al. 1955). These double-cross yellow flint, Venezuela 1; the white Colombian race hybrids were produced by crossing seven inbred lines predominantly from the Cauca Valley, Blanco Comfin; extracted from several populations of Tuxpefio. Two of and a white local variety from near Urrao, Blanco de the collections were from the State of Veracruz, one Urrao. Blanco de Urrao apparently was made up from the State of Coahuila, and one from the State of
primarily of the Colombian races, Comfin and San Luis Potosi. The hybrids were all closely related.
Chococefio. Subsequently, Venezuela 1 and Colombia One inbred (T3) was common in all three hybrids, and 1 were subjected to a host of controlled pollinations two of the three lines were common to two hybrids. The from lines, hybrids, and varieties from Venezuela, hybrids were reported to yield 20% to 25% more than USA, Brazil, Mexico, Puerto Rico, Argentina, and the open-pollinated varieties they were to replace. Cuba. Additional testing, selection, and hybridization Other hybrids were released for the higher elevations of this material led to the selection of 12 ears used to of the Central Plateau. These were based largely on form ETO, a yellow-flint synthetic variety (Chavar- inbred lines from the races Celaya and Bolita (Secreriaga 1966). tariat de Agricultura 1955).
One consequence of the broad genetic background Maize Research Activities in Asia
of ETO was that it had segregating kernel colors. By
selling and test crossing onto a white tester, a large During the 1950s, various other organinumber of pure white ETO lines were selected and zations-USAID, FAO, USDA-also introduced a large recombined to form ETO blanco. Another conse- number of U.S. F1 hybrids into Asia. For the tropical
quence of the diversity within both ETO and ETO areas of Asia, the U.S. hybrids failed to gain a foothold. blanco was that it was possible to develop double-cross First, the U.S. materials were not adapted to the hybrids whose four lines derived entirely from within
one or the other variety. The yield potential of material tropical and subtropical conditions of climate, disfrom these varieties was such that they are still impor- eases, and insect pests. Second, the inbred lines in the tant components of many breeding programs. double crosses were generally too weak to withstand
the harsh production environments of tropical areas.
In the early stages of the Mexican Govern- Third, the yellow dent grain of the U.S. hybrids was not ment-Rockefeller Foundation Program, it was apparent that seed multiplication and distribution of the the preferred grain types. However, the U.S. Corn Belt better landraces might clearly give rapid and substan- dents were used extensively in the temperate areas of tial increases in maize production in the high plateau. Asia; they were also used in subtropical and tropical From 240 collections in the States of Guanajuato, maize improvement programs to take advantage of Jalisco, and Michoacan, 15 superior collections of several desirable traits, though the gene frequency of early, medium, and late-maturing materials were se- the temperate materials was usually less than 50%.

In 1956, the Government of India and the Rockefeller of the world. Many of these were races, improved Foundation signed a memorandum of understanding varieties, or composites of generally identifiable racial that was to have considerable significance for the backgrounds (Wellhausen 1978). improvement of maize in Asia. The agreement pro- The influence of the Colombian and Mexican provided for a cooperative effort in developing a post- grams was particularly strong in Central America, and graduate school at the Indian Agricultural Research to an appreciable degree in the Caribbean Basin. Initial Institute and for cereal improved work on several exchanges of breeding material and informal technical crops, including maize. The maize improvement work cooperation led to increasingly close coordination of of the Rockefeller Foundation-Indian Agricultural breeding efforts and integration of material. ConseProgram (RF-IAP) began in 1957 and made rapid quently, numerous varieties, synthetics, and hybrids
progress. By the early 1960s, the RF-IAP program was were developed and distributed within these regions. receiving many requests from neighboring countries One of several important releases from this area was for genetic stocks and research assistance. Clearly, Tiquisate Golden Yellow, having appreciable germplasm such international assistance was beyond the scope of from Cuba Oriente (personal communication, E.J. IAP. Therefore, a separate organization, the Inter-Asian Wellhausen and E. E. Gerrish). Tiquisate Golden Corn Program (IACP) was established in 1964, with its Yellow was widely grown in Thailand as "Guatemala" headquarters first in India and after 1966 in Thailand. and as "Metro" in Indonesia and the Philippines.
The IACP eventually included 16 countries stretch- Yellow Salvadorefio was developed by mass selection ing from Afghanistan to South Korea and operated in a varietal cross of Yellow Tuxpan and Cuba PD. (MS)7 continuously from 1964 to 1975. Through IACP a large (Primera Reunion Centroamericana 1954). Both these amount of improved germplasm-mainly from the varieties were important in the later development of
Americas-was introduced into the national maize Suwan-1, and Thai Composite #1 (personal communiprograms of Asia during the 1960s. IACP regional cation, E. E. Gerrish). trials and nurseries began in 1969. These uniform yield These exchanges of information and breeding stocks, trials contained improved varieties and promising accompanied by sometimes spectacular increases in breeding materials from participating national pro- yield or improvement of other characteristics, led to an grams. Initially, most of the entries in the regional trials informal network of exchange and collaboration. Most came from India. Later, the Pakistani and Taiwanese often, an improved variety, or hybrid and its compomaize programs supplied outstanding subtropical nent lines, was used on one side of a cross with locally materials, and Thailand and the Philippines supplied adapted material on the other. While many programs outstanding tropical materials (personal communica- were releasing varieties or synthetics, there was ample tions, T. Izuno and B.L. Renfro). opportunity to also release conventional or modified
hybrids when seed production capability was suitable.
Exchanges of Breeding Material and The general trend of all this activity was greater awareInformal Cooperative Networks ness of the genetic background and racial composition
During the 1950s the exchange of breeding mate- of the breeding material.
rial, lines, and synthetics between the joint Rockefeller-Government maize programs in Mexico Emerging Concepts of Germplasm
and Colombia-as well as germplasm exchange with Use and Breeding Methods
maize programs elsewhere-led to identification of Superior inbred lines, single crosses, or improved selected germplasm that combined exceptionally well varieties from one breeding program were often crossed with one another, e.g., ETO with Tuxpefio, and Mon- onto these same categories of breeding materials in tafia with Chalquefio and Tuxpefio. Indeed, the results another area. This was expedient in the early stages of of racial crosses in the Colombian program indicated many programs in order to obtain better material that the race Montafia, including Ecu 573 from Ecua- rapidly for distribution to farmers. This also led to the dor, should combine well with germplasm in Kenya. recognition of certain germplasm as clearly superior to Therefore, Ecu 573 was sent from Colombia to Kenya others. Consequently, superior germplasm was being and crossed with Kitale II. It resulted in greatly in- distributed widely in many areas of the world. It was creased yields and formed the base for continued also becoming apparent that many parts of the world improvement. Similarly, the Mexican program had could not, and should not, sustain a full-fledged hybrid identified superior breeding materials for many parts seed production and distribution program. Without a

reliable supply of quality hybrid seed-and often no Refinements of Population
funds or credit to purchase these materials-hybrid Improvement Methods
maize use spread slowly in most maize-producing Various types of recurrent selection programs at
countries, even though farmers recognized the yield several locations in the Western Hemisphere were superiority of hybrids compared to their own varieties, proving successful for improving maize varieties or
Beginning around the mid-1950s, there was grow- germplasm complexes. Gains were shown for such ing awareness that the development of synthetic varie- traits as grain yield, stronger stalks and root systems, ties from superior germplasm could serve many lowered plant and ear height, greater resistance to less-developed areas of the world. Breeding theory and some diseases and insect pests, certain other physiomethodologies indicated that appreciable improve- logical characteristics, and modifying environmental ment in maize could be attained by various recurrent adaptation. Recurrent selection procedures provided selection schemes. These procedures could be applied many desirable features when applied to diverse to the existing superior germplasm stocks, varieties, germplasm varieties, complexes, or composites. By the races or racial groups, and composites. Here, the 1960s, a number of the breeding programs using these material from each breeding cycle has the potential of techniques on widely grown varieties or on newly being released directly as a variety. This would allow constituted composites (genetic mixtures of varying for wide-scale distribution, even from farmer to farmer, kinds and complexities) were obtaining appreciable without suffering losses in yield or agronomic desira- gains in the traits being selected. A number of outbility, and without having a sophisticated seed industry standing varieties were developed by the RF-Mexican in place. Additionally, the material from each breeding program. Among these were Tuxpefio Crema I, Tuxpefio cycle could also serve as an improved source from La Posta, and Tuxpefio-1 (E. C. Johnson 1974 and which to extract inbred lines for hybrid seed production personal communication; E. J. Wellhausen 1978 and to suit those areas having more advanced agricultural personal communication). capability. These events were conducive to the thought that
The combination of the above events or circum- compositing various germplasm sources such as variestances led to increased understanding about germplasm ties, individual indigenous collections, racial complexes, use. Certain lines or stocks were obviously valuable for or elite breeding materials into new populations or insect or disease resistance, or for other agronomic germplasm pools would be an effective method to characteristics in a particular situation. The outstand- concentrate more desirable and diverse genes into ing races, varieties, or germplasm complexes that fewer numbers of breeding materials. Subsequent contributed greatly to yield were of complex back- population improvement by recurrent selection methground and genealogy. Diversity per se was no guaran- ods would then further concentrate the desired genes. tee in contributing to hybrid vigor or high yield poten- These populations could then be released as improved tial. It was impossible to systematically test all the varieties. Additionally, it was believed that these popumaize collections to determine which would be most lations would serve as better sources of inbred lines (or advantageous to use under the myriad growing condi- at least reduce the number of stocks to be worked on) tions where maize was produced. Gradually, more for hybrids than did the constituent individual stocks effort was put into developing improved populations, used in the original admixtures or composites. Conseas these could be used both in hybrid breeding pro- quently, there was considerable effort made in many grams as well as those concentrating on varietal devel- breeding programs--including by the International Maize opment. Obviously, the important constituents of these and Wheat Improvement Center (CIMMYT) and the programs were those races or germplasm complexes International Institute for Tropical Agriculture already recognized as being exceptional. These in- (IITA)-to form composites from diverse germplasms. eluded ETO, Tuxpefio, Cuban Flint-Cateto, Coastal
Tropical Flints, Celaya, Pepitilla, Chalquefio, Tiqui- ESTABLISHMENT OF sate, and Salvadorefio, to name a few. In addition, INTERNATIONAL MAIZE numerous studies of intervarietal crosses were under- IMPROVEMENT PROGRAMS taken by the Rockefeller Foundation-Government
programs in Mexico and Colombia during the 1950s The successes of the Rockefeller Foundation, and
and 1960s to determine which other races and varieties later the Ford Foundation, in their cooperative agriculwould be most promising. tural programs in various Third World nations, led

"# --_' o- "
Figure 4. A vivid demonstration of different plant heights resulting from experiments in
improving the grain to stover (stalks and leaves) ratio of Tuxpefio Crema I maize. Tlaltizapan Maize Breeding Station, Mexico. Source: CIMMYT.
these two foundations to seek ways to institutionalize international and regional organizations, and private international agricultural research for basic food crops foundations. It supports 13 nonprofit international in developing countries. In maize research, an interme- agricultural research and training centers that focus on diate step in this process was the establishment through food production challenges in the Third World. In Rockefeller Foundation assistance of the Inter-American pursuing their mandates, the IARCs seek to provide Corn Program for Latin America in 1960 and the national programs in developing countries with a range Inter-Asian Corn Program for South and Southeast of products and services, such as improved germplasm, Asia in 1964. With the establishment of the interna- research procedures, training, consultation, and infortional agricultural research centers, the Rockefeller mation. CIMMYT and IITA operate diversified reand Ford Foundations channeled more of their inter- search, training, and information programs designed national food research efforts through these organiza- to support and complement the work of national maize tions and eventually discontinued their support for the research programs. regional cooperative maize programs that had been At the time of their establishment in the 1960s, established in the early 1960s. CIMMYT, IITA, and the International Center for
In 1971, under the sponsorship of the World Bank, Tropical Agriculture (CIAT) all had maize research the United Nations Development Programme (UNDP), programs that were descendents of various Rockefeller and the U.N. Food and Agriculture Organization (FAO), and Ford Foundation bilateral maize research efforts the Consultative Group for International Agricultural with developing country governments. Among the three Research (CGIAR) was formed to support an ex- IARCs, CIMMYT had the largest maize research
panded network of international agricultural research program and considered that it had a worldwide mancenters (IARCs) modeled after the first four. The date for maize improvement. In the early 1970s, CIAT CGIAR today consists of some 40 donor countries, opted to discontinue its maize improvement work and

to rely on CIMMYT to supply improved maize criminately; others were formed systematically on the
germplasm to national programs in Latin America. basis of replicated trials; others on the basis of race, The interest that CIAT did retain in maize was primar- racial relationship, phenotypic characters, knowledge ily concerned with its frequent intercropping associa- of the performance of related materials, and intuition tion with climbing field beans (one of the CIAT man- (personal communication, E.C. Johnson, J.H. Londated crops). In contrast, IITA has expanded its maize quist, and EJ. Wellhausen). By the early 1980s, howimprovement program from an initial focus on the ever, CIMMYT's selection schemes had become more lowland humid tropics of sub-Saharan Africa to in- formalized (Vasal et al. 1982). clude intermediate-altitude regions of tropical Africa. Unfortunately, the exact kinds and amounts of each
In recent years, CIMMYT and IITA have sought to germplasm component used to make the composites improve their research coordination and collaboration or gene pools when the particular strain, line, variety, between their respective maize improvement efforts in or other entity was incorporated-and the precise Africa. In 1980, two CIMMYT maize breeders were methodology employed at each step along the way-are assigned at I1TA to collaborate in the conversion of not known. In some cases, these events could possibly high-yielding varieties adapted to the lowland tropical be determined from field books and the recollections areasytoldtreak-resistantdaersionshandoinaincreasin a the of those involved, but in other cases it is improbable areas to streak-resistant versions, and in increasing the that a satisfactory reconstruction of the methods and streak resistance of Population 43, La Posta, a late, maeilusdcldbmde(ronlomniwhite dent material with good performance in materials used could bemade (personal communicasub-Saharan Africa. In 1985, CIMMYT and IITA also tion, E.C. Johnson, E.J. Wellhausen). Generally in agreed to collaborate in a new joint germplasm im- these matters, CIMMYT's published accounts are too provement effort for the mid-elevation ecologies of vague to be of much value in tracing germplasm flow, eastern and southern Africa, in association with the breeding history, or determining precisely the constituUniversity of Zimbabwe at Harare. (IITA has recently ents of a gene pool or population, except in a broad withdrawn from formal involvement as part of a new overview. agreement with CIMMYT concerning regional divi- CIMMYT's Maize Improvement Program has sion of responsibilities but will continue to be infor- assembled and improved a broad range of germplasm ally involved.) This program is using germplasm complexes. Using a half-sib recurrent selection breedsources developed by IITA, CIMMYT, and various ing methodology and fairly mild selection pressure, national programs. CIMMYT has developed 24 normal maize and 13
quality protein maize (QPM) gene pools, classified
according to zone of adaptation, maturity period, and
grain type and color. (The QPM work has recently
CIMMYT's International Maize been evaluated in NRC 1988.) These gene pools have
Improvement Program been assembled from CIMMY's germplasm bank-with
In CIMMYT's early years, there was a carryover of more than 10,000 collections mostly from Latin personnel, breeding materials, and philosophies from America-and germplasm supplied by maize scientists the Mexican Government-Rockefeller Foundation from other research programs around the world. Using program. But with their new international "mandate" the most promising germplasm in these gene pools and the RF-Mexican maize team in Mexico took on new from other sources, CIMMYT has derived 24 normal research responsibilities, added staff, and expanded and 10 QPM advanced populations that are suited to a the scope of their breeding materials and germplasm range of climatic conditions (tropical, subtropical, development objectives. From this milieu the concept temperate), maturity periods (early, intermediate, late); evolved that CIMMYT would develop composites, grain color (yellow, white); and kernel type (flint, dent, improved populations, and varieties that could be used floury). See table I for details. directly by national programs as varieties or as breed- In a continuous cyclical process, CIMMYT's pools ing sources for hybrids. and populations undergo improvement for yield potenUsing recurrent selection methods, CIMMYT began tial, protein quality, agronomic type, disease and insect forming numerous composites of germplasm with dif- resistance, and tolerance to stresses such as drought ferent grain types and color for use in broad ecological and aluminum toxicity. In Mexico, these materials are zones. Some of the composites were formed indis- screened for resistance to major foliar diseases, stalk

Figure 5. Dr. Surinder K. Vasal, CIMMYT plant breeder, who played a major role in developing quality protein maize. Source: CIMMYT.
and ear rots, and various ear and stalk borers. Addi- involves cooperation with scientists in more than 80 tional disease screening for downy mildew, corn stunt, countries, and each year, over 700 trials are shipped for and maize streak virus is carried out through collabo- testing at several hundred locations. rative research projects with institutions located in Currently, in each IPTT, the fulf-sib progenies of a areas outside Mexico that are more conducive to this particular population are tested at up to six locations work. Improved genetic sources for specific resistances around the world. In a given year, about 15 populations to various diseases and insects and for tolerance to are tested. The results of those trials are used for two certain environmental stresses associated with drought, purposes. First, based upon information provided by cold, heat, and acid soils are also being developed, the trial cooperators, CIMMYT breeders select the
CIMMYT's germplasms are made available to best 50 to 60 families for within-family improvement,
cooperators in national programs through informal recombination, and regeneration of each population sharing of germplasm from either the pools or popula- for the next testing cycle. The second use of the IPTT tions and through the more formalized international results is for the development of experimental varieties maize testing program, which included International (EVs). Site-specific EVs are derived from the best 10 Progeny Testing Trials (IPTTs), Experimental Variety families at each IPTT location. An across-location EV Trials (EVTs), and Elite Experimental Variety Trails is also formed using the 10 best families across all IPTT (ELVTs). This international testing program presently locations. These varieties are advanced to the F2 stage

and dispatched to cooperators in the form of an Experimental Variety Trial (EVT), each of which is evaluated at 30 to 50 locations worldwide. After the data from these trials have been analyzed, CIMMYT Table 1. CIMMYT's advanced normal and quality data from these trials have been analyzed, CIMMYT
protein maize populations tested regularly in selects the top-performing EVs to prepare Elite Exinternational progeny testing trials perimental Variety Trials (ELVTs), which are distributed to from 60 to 80 locations and conducted in much the same way as the EVTs.
Population Although the development of pools, populations,
and open-pollinated varieties will remain a central
21 Tuxpefio-1 TLWD feature of the CIMMYT maize program, in 1985
22 Mezcla Tropical Blanca TLWD CIMMYT established a hybrid development program
23 Blanco Cristalino-1 TIWD in response to growing national program requests for
24 Antigua-Veracruz 181 TLYD research collaboration and assistance. Information is
25 Blanco Cristalino-3 TLWF
26 Mezela Amarilla TIYF being generated about inbreeding depression and
27 Amarillo Cristalino-1 TIYF heterotic patterns of CIMMYT's broad-based gene
28 Amarillo Dentado TLYD pools and populations. IITA has provided CIMMYT
29 Tuxpefio Caribe TLWD with over 500 inbred lines from its hybrid program.
30 Blanco Cristalino-2 TEWF
31 Amarillo Cristalino-2 TEYF The 1980s have also seen an evolution in the breed32 ETO Blanco SIWF ing programs of CIMMYT. More attention is being
33 Amarillo Subtropical SIYF placed on developing new source populations with
34 Blanco Subtropical SLWD resistance and/or tolerance to specific problems: in35 Antigua-Rep. Domin. TIYD
36 Cogollero TIYD sects, diseases, drought stress, and mineral toxicities.
In many cases, CIMMYT's new source populations are 42 ETO-Illinois SLWD being developed using the superior fractions (for a
43 La Posta TLWD specific trait) of CIMMYT standard populations.
44 Am. Early Dent-Tuxpefio SLWD
45 Amarillo Bajio SIYD CIMMYT has also increased its work in germplasm
46 Temp. Amarillo Crist. SEYF development for highland areas (above 1,800 m) and
47 Temp. Blanco Dentado-2 SIWD for intermediate elevations (1,200 to 1,800 m) (in
48 Compuesto de Hugaria TeEID collaboration with the University of Zimbabwe).
49 Blanco Dentado-2 TIWD
61 Early Yellow Flint QPM TEYF IITA's Maize Improvement Program
62 White Flint QPM TLWF
63 Blanco Dentado-1 QPM TLWD The maize improvement program at the Interna64 Blanco Dentado-2 QPM TLWD tional Institute for Tropical Agriculture (IITA) in
65 Yellow Flint QPM TLYF Ibadan, Nigeria, was initiated in the early 1970s. Its
66 Yellow Dent QPM TLYD major objective has been to develop high-yielding
67 Temp. Blanco Grist. OPM SLWD
68 Temp Blanco Dent. QPM SIWD germplasm with increased yield dependability by
69 Templado Amarillo QPM SIYD emphasizing resistance breeding through recurrent
70 Temp. Amarillo Dent. QPM SIYD selection. Initially, efforts were devoted mainly to
developing improved germplasm with resistance to Puccinia polysora and Helminthosporium maydis for the lowland humid tropics. Two late-maturing white Key populations were developed initially: TZB (African
T = Tropical L = Late maturity and Latin American sources with Nigerian Composite
S = Subtropical I = Intermediate maturity B as the most important component) and TZPB (deTe = Temperate E = Early maturity rived from CIMMYT's Tuxpefio Planta Baja Cycle 7).
F = Flint QPM = Quality protein maize Varieties developed from these two populations have
D = Dent Y = Yellow been released in Nigeria, Cameroon, Gabon, and Benin.
In 1975, work was initiated to develop germplasm
with resistance to Maize Streak Virus (MSV), a disease currently confined to the African continent. This re15

After the discovery of MSV resistance, an intensive breeding program was initiated to develop resistant germplasm (populations, varieties, and inbred lines) with different ecological adaptation, maturity, grain color, and texture. TZSR-W and TZSR-Y, late white and yellow semiflint populations adapted to the lowland tropics, respectively, were the first improved populations with MSV resistance developed at IITA. Since 1980, CIMMYT has also posted one or two maize breeders at Ibadan to work with IITA scientists in the conversion of additional germplasm to streak-resistant types. In addition to developing new streak-resistant populations to serve additional ecologies, the CIMMYT-ITA collaborative program has converted, through backcrossing, over 30 experimental varieties from more than a dozen existing CIMMYT populations that are well-adapted to African maize production conditions. Streak resistance is also being incorporated into some of the most widely grown commercial varieties from African countries through a backcrossing conversion program. Today, many streak-resistant maize materials are being made available to national programs through IITA's international testing program.
In 1979, IITA initiated a tropical hybrid development program to produce adapted, disease-resistant inbred lines with good combining ability for use by Figure 6. Testing of IITA inbred lines, for use in national maize programs in hybrid and synthetic varideveloping hybrid varieties, at a savanna station in ety development. The hybrid program was expanded in Nigeria. Source: CIMMYT. 1982 through a special project funded by the Government of Nigeria; it subsequently evolved into a joint IITA-Nigeria hybrid program. IITA is now giving more search effort has been the major aim of the IITA maize attention to transferring hybrid technology to other serceffog rt hasung the ajo caim of th A maie interested African countries such as Cameroon, Ghana, program during the last decade (IITA 1986a). MSV is adCt 'vie one of the most economically important diseases in Africa. It occurs over a wide range of ecologies, from A number of outstanding hybrids for tropical areas sea level to 1,800 m elevation and from the humid have been developed through this research effort (IITA forest zone to the dry savanna. Yield losses of up to 1986b). These materials are made available to national 100% due to MSV have been recorded in manyAfrican maize researchers through two international hybrid countries. Because of the erratic natural occurrence of trials (white and yellow) that are distributed throughMSV, it was necessary to develop an artificial out Africa and in a growing number of countries of mass-rearing facility for the vector, Cicadulina spp., Asia and Latin America. Commercial production of and techniques for uniform infestation in the field. these hybrids is already under way in Nigeria. Through this screening program, two sources of resis- In addition to its work in hybrid development and tance were identified: the variety TZ-Y from IITA and the breeding for MSV resistance, IITA's maize imthe variety La Revolution developed by the French provement program has also concentrated on five Tropical Agricultural Research Institute (IRAT) at other priority research areas: Reunion Island. These two sources have been used to 1. Combined resistance to MSV and downy mildew. develop multigenic resistance to MSV. While resistant While downy mildew, Peronosclerospora spp., is not plants are not immune, they show reduced symptoms widespread in Africa, areas infested with downy milof MSV. dew generally overlap those subject to streak virus

attack. White and yellow populations with combined different germplasm sources indicating that breeding resistance to the two diseases have been developed for Striga resistance is possible. A number of sites with using downy mildew resistant sources from Thailand a high level of natural infestation have been identified and the Philippines. These materials have been crossed as locations for screening. Work has also begun to with streak-resistant adapted materials. The level of develop laboratory techniques for artificial infestation. resistance to the two diseases is high, and preliminary 5. Breeding for the mid-altitudes. The mid-altitude experiments have shown that the resistant materials ecologies of sub-Saharan Africa have some of highest have good yield potential. Inbred lines and hybrids with yield potential. Germplasm requirements for these combined resistance to MSV and downy mildew are zones included resistance to leaf blight (H. turcicum), being developed; some of these materials are currently rust (P. sorghi) and MSV. In 1978, IITA began to being used in Thailand and the Philippines (personal develop maize populations with resistance to these communication, S.K. Kim). diseases, and these populations have been tested in 20
2. Resistance to stem borers. Work is in progress on countries. In 1985, CIMMYT, IITA, and the University resistance to the two major stem borers species found of Zimbabwe, Harare, launched a collaborative in West Africa: Sesamia calamistis and Eldana sac- mid-altitude germplasm development program for charina. Development of suitable artificial insect rear- maize-growing areas in eastern and southern Africa ing and infestation techniques is still under way and between 1,200 and 1,800 m elevation. Improved little research impact has been achieved to date. populations, inbred lines, and open-pollinated varie3. Drought tolerance. Breeding for drought toler- ties with enhanced disease resistance for mid-altitude ance was initiated in 1982 in Burkina Faso through the ecologies are being developed through this joint proSAFGRAD/IITA collaborative project. (CIMMYT gram. (As noted earlier, IITA has recently withdrawn
also participates in this work.) Breeding for early- and from formal involvement as part of a new agreement extra-early-maturing germplasm is emphasized, with with CIMMYT concerning regional division of responsagronomic research also being an important compo- ibilities but will continue to be informally involved.) nent. Issues of water conservation, soil fertility, and soil While the programs outlined in this chapter are, in compaction are the major topics in the agronomy the first instance, designed to help developing nations, research program. the genetic products may also prove to be of value in
4. Resistance to Striga spp. This parasitic weed is breeding programs in developed nations. One exbecoming a serious problem in the expanding ample of the potential is provided in a recent article by maize-growing areas of the African savanna. A good Holley and Goodman. (The abstract is reproduced in level of resistance has already been identified among appendix B.)

This chapter provides a preliminary status report of early-maturing and/or drought-tolerant materials, (2) the recent releases and uses of open-pollinated varie- areas with additional rainfall requiring ties and hybrids in 57 developing countries. The coun- intermediate-maturity materials, and (3) late-maturing try reports probably overemphasize the germplasm materials suitable for areas with a long growing season contributions of CIMMYT, IUTA, and predecessor and ample rainfall. Breeding efforts to meet the needs international programs, such as those of the Rockefeller of these different moisture regimes include the develFoundation. This is primarily because more informa- opment of open-pollinated varieties as well as hybrids tion is readily available from these international or- with resistance to H. turcicum and P. sorghi. An ganizations than from national programs. It is hoped early-maturing, open-pollinated variety developed from that future efforts to trace the use and impact of this work was submitted in 1985 to the National Variety improved maize genotypes will help to fill in the infor- Release Committee for final approval. mational gaps that are contained in this report. A national seed program was implemented in 1978 with the establishment of the Ethiopian Seed CorporaSUB-SAHARAN AFRICA tion (ESC). Large-scale seed production and distribution has been carried out since 1980. In 1985, the ESC
Eastern and Southern Africa reported seed sales of 11.8 t; of this total, 7.5 t was
commercial seed, not certified but field approved.
IITA reports supplying the ESC with2 t of TZSR-1 and
ETHIOPIA 200 kg of TZESR-W for multiplication and distribuMaize is produced throughout Ethiopia in temper- tion to farmers in a new settlement area where maize ate as well as tropical environments, with the bulk of streak virus is widely present (personal communicaproduction concentrated in the southern, southwest- tion, Y. Efron). The ESC also has initiated production ern, and western regions. During 1983-85, approxi- of two hybrids (8321-18 and 8322-13) developed at mately 1.2 million t of grain were produced annually on IITA. 850,000 ha, with national yields averaging 1.4 t/ha.
Among the major cereals, maize ranks first in total KENYA production and in yield; it ranks fourth in total area Maize is the major staple for Kenya's population, behind Tef (Eragrostis abyssinica), barley, and sor- and more than 90% of this maize is produced by ghum. Almost all of the maize produced is used for small-scale farmers. During 1983-85, approximately human consumption. 2.4 million t of grain were produced annually on 1.5
Maize research is conducted by the Institute of million ha, with national yields averaging 1.5 t/ha. In Agricultural Research (LAR) and by the College of addition, some 100,000 t of maize were imported annuAgriculture (Debelo 1985). Both CIMMYT and IITA ally during 1982-84. report collaboration with national scientists. Germplasm Maize research in Kenya began in colonial times development efforts are directed toward the produc- and has continued to evolve since independence. tion requirements of three major zones: (1) areas with Germplasm development is directed toward four maa short rainy season or erratic rainfall requiring jor agroecological zones (Ochieng 1985):

Figure 7. Ear of hybrid maize developed in Kenya. Source: A.I.D.
1. High-potential zone, unimodal rainfall pattern als were carried out and the first hybrids were released. (1,000 to 2,200 mm), 1,600 to 2,300 m altitude Several improved varieties and conventional hybrids
2. Medium-potential zone, bimodal rainfall pattern were made available to farmers: Kitale II and hybrids (700 to 1,800 mm), 1,000 to 1,700 m altitude H 631, H 621, and H 622. Also, a variety cross H 611
3. Low-potential zone, scanty, short-duration rains (KII x Ecu 573) was released for commercial use.
4. Coastal strip, hot, humid belt, some saline soils A maize breeding program was undertaken at During the 1950s, local maize landraces-tracing to Kitale, Kenya, as part of the U.S. cooperation with the probable affinities with Tuxpeflo-Hickory King taken East African Agriculture and Forestry Research Orto southern Africa in the 19th century, then selected ganization (EAAFRO). To strengthen this cereal over generations for adaptation as the crop slowly improvement research, the U.S. Agency for Internamoved north-were gathered from farmers' fields. These tional Development and the U.S. Department of Agrimaterials were used to develop composite populations culture agreed in 1963 to cooperate in a Major Cereals to form the basic breeding stocks. However, because Project in Africa, which included Kenya (Eberhart and the germplasm base had been markedly narrowed Sprague 1973; Johnson et al. 1980). The USAID-USDA
through selection, efforts to improve this material met Program set up a comprehensive breeding system with little success, and local breeders looked for new using two composites (Comp. B and Comp. R) which sources of genetic diversity, were improved by recurrent selection.
In the late 1950s, new exotic germplasm, notably Since 1963, more than a dozen open-pollinated Ecuador 573 and Costa Rica 76, was introduced into varieties and composites and various types of hybrids the breeding stocks, and a search began for maizes with have been released for commercial production (table heterosis in crosses with local strains. During the 2). Hybrid maize has become so popular in Kenya that 1960s, inbreeding and hybridization with local materi- most farmers in the more favored production areas will

Table 2. Maize varieties released by IITA (personal communication, S.K. Kim). Breeding
Kenyan National Breeding Programs materials from the CIMMYT/University of Zimbabwe
collaborative mid-altitude breeding program are also
Year of Yield being supplied from Harare.
Varietya Typeb release (% of KSM) The Kenya Seed Company (KSC), in which the
KSM OP 100 Government of Kenya now has 51% ownership, has a
ECU 573 OP 1959 monopoly on seed production in the country and also
KS II OP 1961 107 exports seed (hybrids) to neighboring countries. Maize
H 611 VC 1964 142 seed constitutes its largest crop, with a volume in 1984
H 621 DC 1964 132 of 14,000 t. In 1985-86, the KSC produced 10 hybrids
H 631 TWC 1964 140 and 2 open-pollinated varieties for the various agrocliH 622 DC 1965 135 matic areas of the country. Breeder seed for these
H 632 TWC 1965 140 varieties comes from three agricultural research staH 612 TC 1966 155 tions and from KSC's small but effective maize reKCB OP 1967 search program. Most farmers in the high potential
H 511 VC 1967 areas use improved seed. In marginal regions, lanH 512 VC 1970 draces are still often found.
H 611C VC 1971 155
H 613 TC 1972 166 LESOTHO
CMC OP 1974 Maize is the staple food in Lesotho. In 1983,90,000
H 614 TC 1976 166 t of grain were produced on 60,000 ha, with national
H625 DC 1981 176 yield averaging 1.5 t/ha.Lesotho is entirely surrounded
Source: Ochieng 1985 by South Africa. Maize growing areas can be divided
aKCB = Katumaini Composite B; CMC = Coastal into three major agroecological zones: the lowlands, Maize Composite; the foothills, and the Maloti mountain range, where
bop = open-pollinated variety; VC = variety cross; land is cultivated as high as 3,000 m elevation.
DC = double-cross hybrid; TWC = three- way cross Maize research is carried out by Agricultural Rehybrid; TC = top-cross hybrid search Lesotho (Ntlhabo and Math 1985). For the low
elevations, South African hybrids are the dominant
materials grown. In the more mountainous regions,
not accept anything else. Between 1963 and 1981, the because of the shorter growing season and cooler hybrid area has increased from zero to nearly 600,000 conditions and the subsistence nature of maize producha. Until the late 1960s, large-scale farmers dominated tion, open-pollinated varieties are more suitable genothe use ofhybrids; since then, the number of small-scale types. CIMMYT is collaborating with Lesotho refarmers has increased greatly. searchers to develop high-yielding varieties for the
Maize materials from both CIMMYT and IITA highland areas. Pools 1, 2, and 4 supplied by CIMMYT continue to be evaluated by national maize program markedly outyield the two best South African hybrids, scientists. Some have been used for the development of SA4 and SAll, in this zone. Seeds from these materials several experimental hybrids yet to be released. A- have been multiplied and are being tested extensively though CIMMYT populations and pools are generally through on-farm trials. CIMMYT reports that Pools 1 susceptible to common rust and turcicum blight, some and 4 have been released as open-pollinated varieties are being used in the breeding program. CIMMYT and are under commercial production. Populations 21 (Tuxpefio-1), 49 (Blanco Dentado-2),
and 32 (ETO Blanco) have been used in the development of several varietal hybrids in Kenya: PWAN-1, Maize is the principal staple food in Malawi. DurH-i, H-2, and H-3. ing 1983-85, approximately 1.4 million t of grain were
A good streak virus tolerant composite (TZSR) produced annually on 1.2 million ha, with national from IITA has been included in the B population in yields averaging 1.2 t/ha. In recent years, Malawi has Kenya. Streak-resistant inbred lines for both lowland been a net maize exporter. Maize is grown by small-scale and mid-altitude environments have been supplied by farmers for food and cash and by large-scale farmers

Table 3. Seed produced by the National Seed Company of Malawi, 1978-79 to 1983-84, in tons
Material 1978-79 1979-80 1980-81 1981-82 1982-83 1983-84
MH 12 (single
cross) 911 1,143 2,123 1,704 1,617 3,015
MH 13 (3-way
cross) 37 166
NSCM 41 (3-way
cross) 9 236 574
UCA Basic
(composite) 6
UCA (certified) 201 515 2,231 153 135 333
CCA Basic
(composite) 7 70
CCA Certified 90 35 44 55 110 53
A 73N (inbred line) 42 19 52 57 1
B 76S (inbred line) 8 5
Source: National Seed Company of Malawi
mainly for commercial purposes. There has been a adaptability. A number of the newly released hybrids in remarkable increase in the area grown with maize over Malawi, mostly single-cross hybrids, have been derived the past five years, though not in yields. from these lines.
About one-quarter of the maize area in Malawi is CIMMYT materials-especially white flints-have planted with improved materials: 17% with improved served the national program in two ways: (a) for direct open-pollinated maize (composites) and 9% with hybrids or indirect use in the formation of open-pollinated (personal communication, A. Manwiller). The local varieties for the small-scale farmer and (b) to form new unimproved varieties are mostly flints to semi-flints, source populations. An open-pollinated variety, Chitwhich are preferred because of their resistance to edze composite C, was released in 1976. This composstored-grain insect attack and because of their ite was formed by crossing 36 entries selected on the flour-making qualities. Initially, maize breeders were basis of their visual performance in two CIMMYT using dents in their improvement programs. CIMMYT
has supplied improved flint germplasm, and these variety trials, EVT 12 and ELVT 18, consisting mostly materials are being used in crosses with Malawian of late tropical materials of variable kernel types. dents. Hybrid maize is grown only for commercial Tuxpefio 1 (Pop. 21) was also identified from the purposes where industrial millers sell meal mainly to CIMMYT experimental varieties as being well adapted the urban population. to the northern lakeshore area of Malawi. This variety
Maize improvement work in Malawi is focused on was released in 1985 under the same name.
production problems in two ecological zones: (a) high Since 1981, the Malawi Improvement Program has potential areas where rainfall is not a limiting factor been cooperating with IITA. It has received streak(more than 600 mm) to successful maize growth and resistant materials that have been incorporated into
(b) marginal areas where rainfall (less than 600 mm) is national breeding populations for evaluation under a limiting factor (Ngwira and Sibale 1985). The pro- local conditions of natural infection. Inbred lines are gram for marginal areas is evaluating materials with being extracted from one population, TZESR, which is earlier maturity from CIMMYT, IITA, and South moderately resistant to streak virus attack. IITA inbred Africa. The materials obtained from South Africa are lines and hybrids have also been supplied to maize mainly inbred lines that perform better than materials researchers at the Chitedze Station (personal commufrom other regions of the world in terms of disease and nication, S.K. Kim).

The National Seed Company of Malawi (NSCM), REUNION established in 1976, multiplies the seed of improved
materials by contract with commercial farmers. Seed The Department Cultures Vivrieres Du Cirad of both hybrids and open-pollinated varieties has been cooperates with IITA and CIMMYT in germplasm produced (table 3). exchange and testing (Marchand and Hainzelin 1985).
At present, NSCM is handling two hybrids: MH12, The breeding program is devoting major efforts to the a single-cross hybrid, and NSCM 41, a three-way cross study of four viruses: maize streak virus, maize mosaic hybrid. Parent seed for two newly released hybrids was virus, maize stripe virus, and sugar cane mosaic virus. handed over to the company in 1984, and production of Since vector insects are important in virus transmission the new hybrids was planned to commence in the 1985- in maize the entomological aspects of these diseases 86 cropping season. NSCM also produces seed of are also under study. IITA's streak-resistant materials composite varieties. At present, it is producing Ukirigula are also being utilized. IITA's inbred lines are also used Composite A (UCA) and Chitedze Composite A (CCA). to test strain variations of maize streak virus (personal Tuxpefio 1 will shortly be added to the list of communication, S.K. Kim). Tocumen (1) 7931, seopen-pollinated maize seed multiplied by the com- elected from CIMMYT material in 1984, is being inpany. creased despite susceptibility to H. turcicum. Most
farmers grow open-pollinated varieties, but one hyMOZAMBIQUE brid, IRAT 143, has had some success.
Maize is the main staple food of most Mozambi- Seed production in Reunion is very limited and not Maizd is well organized. Scattered private institutions produce cans and is grown throughout the country. During seed but there is no organized agency. In 1985, im1983-85,333,000 t of grain were produced annually on proved seed of the following genotypes was produced 600,000 ha, with national yields-the lowest in (Marchand and Hainzelin 1985): Revolution (local sub-SaharanAfrica-averaging 0.6 t/ha. Disease prob- improved variety), 5 t; IRAT 143 (hybrid), 1 t; and lems (mainly downy mildew and maize streak virus), Tocumen (1) 7931, 5 t. insect pests, weeds, and lack of water are all serious
biological constraints in one or more of the three major SOUTH AFRICA agroecological zones in which maize is grown. Maize is the major staple for most of the country's
Maize research is carried out by the National people. During 1983-85, approximately 5.3 million t of Agricultural Research Institute (INIA), which initi- grain were produced on 4.1 million ha, with national ated the national maize research program after inde- yields averaging 1.3 t/ha. During 1982 and 1983, severe pendence in 1977 (Nunes and Sataric 1985). In recent drought reduced national production: viz, 13 million t years, social and political upheavals have seriously in 1980 and 4 million t in 1982. In most years, however, affected maize research efforts, and many valuable South Africa is an important maize exporter. materials have been lost. Almost all of the South African maize area is
Several open-pollinated varieties based on CIM- planted with hybrids--three-way and four-way crosses MYT materials have been selected for release. Foun- (personal communication, Mike Barrow). Most of dation seed has been produced for five experimental these hybrids are based on material from the United varieties: Obregon 7643, Cotaxla 7921, and Ferke 7822 States that was introduced during the early part of this for high altitudes; and San Andres 7823 and Mexico century (personal communication, D. Duvick). There 8049 for the northern lowlands. Streak resistant mate- are no public sector seed companies handling maize rials from IITA have also been obtained and are being seed. Seven private companies are active in South used in the breeding program; a variety based on Africa. Materials from CIMMYT and IITA have not TZESR-W has been selected for commercial releaIse been well adapted, but they are screened for disease (personal communication, S.K Kim). Through techni- resistance and for resistance/tolerance to African Maize cal assistance from the Yugoslavian Maize Institute, a Stalk borer, Busseola fusca. The Pioneer Seed Comhybrid development program was started in 1984. pany (Pty) Limited (no association with Pioneer Hi-Bred
Three national institutions are involved in maize International Co. of the United States) has received seed production: INIA is responsible for producing open-pollinated populations from both IITA and breeders' and basic seed; the National Seed Company CIMMYT. Inbred lines are being developed from the (ENS) is responsible for certified seed production; and most promising materials (80% to 85% of effort), and the National Seed Service (SNC) is charged with qual- population improvement makes up the remainder of ity control. the breeding program.

borer Chilopartellus. CIMMYT, IITA, and the SemiArid Food Grains Research and Development project (SAFGRAD) are cooperating with CARS in these breeding efforts. Nearly 200 experimental varieties supplied by CIMMYT have been evaluated in field trials since 1981. Promising materials include Across 8121, Across 8149, Los Diamantes 7823, Across 7822, Pirsabak 7930 (early maturing) and Poza Rica 7926. The best performing materials from the SAFGRAD trials have been Pool 16 (early maturing) and TZPB. Pirsabak 7930 and Pool 16 are being crossed with the F1 progeny of the locally developed variety-cross hybrid, ISOMA, in order to introduce greater earliness into this material.
Maize is the principal staple in Swaziland. Approximately 90,000 t of grain were produced annually in 1983-85 on 60,000 ha, with national yields averaging 1.5 t/ha. Maize production has been cyclic depending on climatic conditions in recent years.
Approximately 80% of the land is planted with hybrids, with the balance planted with local open-pollinated materials or second-generation hybrid seed (Shikhulu and Mavimbela 1985). The main hybrids in use are SR 52 (from Zimbabwe) and NPP x K64R (from South Africa). Seed of hybrid maize is Figure 8. CIMMYT maize trainee from Africa. produced and distributed by a national agency, the
Source: CIMMYT. Swaziland Seed Multiplication Project. Several private
companies and cooperatives also distribute maize seed.
Inbred lines and varieties with maize streak virus SOMALIA resistance are received from IITA. The program also
receives open-pollinated varieties from CIMMYT that Maize is the second most important crop in Somalia, are evaluated for yield and general adaptation. One after sorghum. During 1983-85, approximately 130,000 such variety, Across 7443, is now commercially availt of grain were produced annually on 138,000 ha, with able. Several more varieties-based on CIMMYT national yields averaging 0.9 t/ha. Populations 22 (Mezcla Tropical Blanca), 23 (Blanco
Maize research began in 1977 and is carried out by Cristalino-1), 24 (Antigua-Veracruz 181), and 47 the Central Agricultural Research Station (CARS) at (Templado Blanco Dentado-2)-have been selected Afgoi (Abbanur 1985). In 1979, the variety Afgoi for further testing and possible eventual release. Composite was developed from Somali landraces, Guatemala Flint, and U.S. hybrids. Another variety, TANZANA Somtux, was developed in 1980 from half-sib crosses between Afgoi Composite and Tuxpefio (obtained Maize is the most important food crop in Tanzania.
from Tanzania). Somtux is a white, semi-dent grain, In 1983-85, approximately 1.6 million t of grain were full-season variety. Because of discontinuity in the produced annually on 1.5 million ha, with national breeding programs, both varieties have lost their ge- yields averaging 1.1 t/ha. Most farmers prefer white, netic purity and have become contaminated. flint-type maizes.
The Somali maize program's germplasm improve- Tanzania's National Maize Research Program
ment priorities are to develop high-yielding varieties (NMRP) was established in 1973 with the assistance of with tolerance to drought and resistance to the stem IITA, CIMMYT, and USAID. Maize research is fo24

cused on three agroecological zones (personal corn- Table 4. Seed production of improved
munication, AJ. Moshi): (1) the lowland zone, includ- maize genotypes in Tanzania, 1972-73 to ing coastal and other areas below 900 m elevation, (2) 1983-84, in tons
a mid-altitude zone between 900 and 1,500 m but
subdivided into two sub-zones: (a) areas with more Openthan 1,100 mm rainfall and with a longer growing pollinated
season, and (b) areas with less than 1,100 mm rainfall Season Hybrids varieties Total
and with a shorter growing season, and (3) the highland 1972-73 420 1 421
zone above 1,500 m elevation-the major production
area-which usually receives sufficient rainfall and has 1973-74 666 109 775
relatively long growing seasons. Maize streak virus is a 1974-75 1,366 1,050 2,416
problem in the lowland and mid-altitude zones, and H.
turcicum, ear rot, and stalk borers can be serious
problems in the high-altitude zones. 1976-77 916 2,128 3,044
The breeding program has been involved in ira- 1977-78 409 1,061 1,470
proving populations, in the formation and testing of 1978-79 2,485 1,615 4,100
varieties, and in the supplying of national foundation
seed farms with breeders' seed of both open-pollinated 1979-80 3,022 107 3,129
varieties and inbred lines. A major effort is aimed at 1980-81 2,129 1,516 3,645
formation of open-pollinated varieties for commercial 1981-82 1,525 851 2,376
use by farmers. About 10% of the breeding program
involves inbred development and hybrids, but this 1982-83 1,909 1,465 3,374
activity is increasing. Beginning in the 1982-83 season, 1983-84 2,537 1,114 3,651
the NMRP began forming top crosses for possible Source: Personal communication, A.J. Moshi.
hybrid production and for obtaining information for
restructuring the populations on the basis of heterotic
The NMRP of Tanzania cooperates with various the intermediate-elevation areas. For several years, international breeding programs including CIMMYT, Illonga Composite has been recommended for areas IITA, and SAFGRAD (personal communication, AJ. below 900 m elevation. The full-season variety, Tuxpefio Moshi). International progeny trials are tested for local (Population 21, Tuxpefio-1) is grown on a limited scale adaptation; variety trials are received for both in the northeastern and southern lowlands. Until reopen-pollinated and hybrid materials; and seed is re- cently the variety Katumani was the only early-maturing quested of varieties, inbreds, and breeding material, material available. In 1983, the NMRP released a new The materials are either tested locally or used in early-maturing variety, Kito (Population 30, Blanco breeding and selection for local improvement. Four Cristalino-2), for areas below 1,300 m elevation and open-pollinated varieties based on CIMMYT germplasm two full-season varieties, Kilima (Pop. 21) and Staha have been released for commercial production. Mate- (Pop. 21), for the intermediate and low-altitude zones, rials from IlTA are evaluated for streak virus diseases respectively. resistance. TZESR-W has shown high resistance to Improved variety and hybrid breeder's seed is inMSV and good yield potential in national trials (per- creased to foundation seed at national foundation seed sonal communication, S.K. Kim). None of the IITA
material has been released, but some of the streak farms T naniSeedoComany esale in
resistant converted materials and newly tested hybrids 1973, contracts with institutions and larger scale farmare promising. ers to produce certified seed. The latter is marketed
Presently, two hybrids, H 6302 and H 614, devel- solely by the Tanzania Seed Company. There is no oped by the East African Agriculture and Forestry private maize seed company. The use of improved seed Research Organization (EAAFRO), are recommended has increased significantly since 1973 (table 4). Even for the high-elevation, long-season areas of the south- so, many farmers still do not buy certified seed because ern highlands. The hybrids, H 632 and H 622, and the of cost factors, delivery problems, and a lack of knowlopen-pollinated variety, UCA, are recommended for edge about the importance of using quality seed.

UGANDA ect have screened over 200 streak-resistant lines in a
national hybrid development program. A collaborative
Maize is an important staple in Uganda. During Zambian/Yugoslavian maize program has also used 1983-85, approximately 365,000 t of grain were pro- alin slfr an me poams asisn duced annually on 358,000 ha, With national yields II1TA lines for hybrid development. IRTAs streak-resistant averaging 1.2 t/ha. mid-altitude population (TZMSR-W) and several of
ve Ugagi n aze pthe CIMMYT/IITA streak-resistant conversions of Theexperimental varieties show promise in Zambia (perby internal war and must be rebuilt. Cooperation with sonal communication, S.K. Kim). IITA and CIMMYT will be very important in supplying The Zambia Seed Company (Zamseed) has been populations and inbred lines (personal communica- responsible for the production, procurement, and distion, Oguti W. Mangheni). IITA has supplied tribution of maize seed. The seed is produced by streak-resistant material for the lowland areas and has trrbution ohe Seed reduce d by been assisting in converting local varieties to streak are int Zambin Seed oto resistance. The Ugandan maize program is also inter- and qualityaconroli uted he Seed Cntrol ested in releasing varieties from CIMMYT Popula- and Certification Institute. The main hybrid in use is etdins rTelesiog 22rieezc Tro pc BIPpulana ~ SR 52 (single-cross) which was developed in Zimtions 21 (Tuxpefio-1), 22 (Mezcla Tropical Blanca, 43 babwe. In 1983, an improved version of SR 52, named (La Posta), and 49 (Blanco Dentado-2) (Kaahwa and bw I1, an impoved veson of Sn Kabeere 1985). MM 752, was released and has shown great promise. In
1984, seven other hybrids were released: MM 501, MM
The most popular improved open-pollinated varie- 504, MM 602, MM 603, MM 604, and MM 606. In ties in Uganda are Kawanda Composite A and Katu- addition, two open-pollinated varieties based on mani, a Kenyanvariety used in the semi-arid Karamoja CIMMYT materials, MMV 400 (Pirsabak (2) 7930, region. Some hybrid seed is also imported from Kenya, Pop. 30, Blanco Cristalino-1) and MMV 600 (Pop. 21, mainly H 622, H 632, and H 614. The Uganda Seed Tuxpefio-1), were also released and are being extenProject (USP) is responsible for the production of sively grown. These new cultivars provide Zambian improved seed, which is marketed by farmers' coop- maize producers with the needed flexibility to adjust eratives as well as by the USP. The Karamoja Seed planting dates according to the onset and duration of Scheme also produces improved seed for farmers in the rainy season (Chibasa 1985). that region.
ZAMBIA Maize is the principal staple in Zimbabwe. During
Maize is the most important crop in Zambia. It is 1983-85, approximately 1.7 million t of grain were grown in most regions of the country, except in excep- produced annually on 1.3 million ha, with national tionally dry, wet, or infertile areas where sorghum, yields averaging 1.2 t/ha. Both total production and millet, or cassava are better adapted. During 1983-85, yields during this 2-year period were 25% below averapproximately 863,000 t of grain were produced annu- age due to severe drought. In most years, Zimbabwe is approximately ,000 t ofith rainl wered aerin a net maize exporter. ally on 512,000 ha, with national yields averaging 1.7 t/ Zimbabwe's maize breeding program was started ha, among the highest in sub-Saharan Africa. in 1932 and is one of the oldest in Africa. The initial
Hybrids dominate commercial production areas, genotypes developed were open-pollinated varieties. and open-pollinated varieties dominate subsistence In 1950, Southern Cross was the dominant variety agriculture areas. Approximately 30% of the total grown in the country. Since then, 14 double, 5 three-way, maize area is planted with hybrids and 15% with 6 single, and 6 modified single-cross hybrids have been improved open-pollinated varieties and composites released (personal communication, R.C. Olver). The (Chibasa 1985). Maize streak virus can cause serious single-cross hybrid, SR 52, released in 1960, replaced economic losses in certain years (personal communi- SR 14 because it was higher yielding. For more than cation, S.K. Kim). two decades, SR 52 has continued to be the dominant
CIMMYT and IITA collaborate with Zambian hybrid in Zimbabwe (grown on 85% of hybrid area)
maize researchers, and international trials from both and is also widely grown in other African countries IARCs are regularly received and grown. Two (Billing 1985). Two shorter season hybrids, R 200 and
open-pollinated varieties based on CIMMYT germplasm R 201, are grown on the remaining 15% of the hybrid have been released by the national program. Since maize area. R 201 has tended to replace R 200, which 1980, scientists in a joint Zambia/FAO research proj- is susceptible to weevil damage (Billing 1985).

Figure 9. Pounding maize to make corn meal for home use, Africa. Source: CIMMYT.
The Zimbabwe maize breeding program cooper- high resistance to MSV and H. turcicum (personal ates with IITA and CIMMYT. This collaboration was communication, S.K. Kim). CIMMYT's international strengthened substantially with the establishment of a progeny and variety trials are routinely evaluated; joint mid-altitude breeding station in cooperation with Population 48 is also being used to reduce plant height the University of Zimbabwe. The station was estab- in various national materials. Relatively small proporlished jointly by CIMMYT and IITA in 1985 and tions of CIMMYT and IITA germplasm have been operated for 2 years under joint sponsorship. IITA has introgressed into a few of the 28 locally constituted recently withdrawn from formal involvement as part of populations, which are undergoing various forms of a new agreement with CIMMYT concerning regional recurrent selection to serve as long-term genetic sources division of responsibility (IITA for western and central of new inbred lines. None of the released hybrids have Africa, and CIMMYT for eastern and southern Af- any germplasm from either CIMMYT or IITA. rica). But IITA will continue to be informally involved
and cooperate with CIMMYT. The Seed Co-op Company of Zimbabwe, a coopThe greatest contribution from the IARCs has erative of commercial farmers, has the sole right to been the identification of streak virus resistant germplasm market government-developed hybrids. All commerand the incorporation of this into different local popu- cial farmers grow hybrids and at least 70% of the lations. UTA's mid-altitude population, TZMSR-W--and communal peasant farmers also grow hybrids (Billing the inbreds and hybrids derived from it-have shown 1985). The commercial farmers grow about 250,000 ha

of maize and the communal farmers plant about 1,000,000 improvement for lowland, mid-altitude, and highland ha. Commercial yields generally average 5 to 6 t/ha ecologies. In view of the importance of maize streak while the communal group yields average between 1 virus in the lowland areas, only streak-resistant varieand 2 t/ha. Since Independence, the Zimbabwe Gov- ties are being released. Several streak-resistant varieernment has given priority to increasing yields among ties based on IITA populations and the CIMMYT/ the communal farmers: considerable headway in trans- IITA conversion program have been released: TZB ferring improved varieties and production technolo- and TZPB are currently grown on approximately 10,000 gies has been made during the 1980s. ha. Since 1984, IITA hybrids have been tested extensively; some of these hybrids are showing tolerance to
Western and Central Africa Striga (personal communication, S.K. Kim).
Maize is the major food staple in Benin. During Maize is an important food crop in the Central 1983-85, approximately 370,000 t of grain were pro- African Republic. In 1983, approximately 30,000 t of duced annually on 485,000 ha, with national yields grain were produced annually on 100,000 ha, with the averaging 0.7 t/ha. average national yield equal to 0.3 t/ha, the lowest in
The national maize program cooperates with sub-Saharan Africa.
CIMMYT, IITA, and SAFGRAD. The variety TZB Tuxpefio-1 was introduced in the early 1980s and is
received from IITA is grown in the northern region; an being used. International variety trials were supplied experimental variety from TZSR-W-1 is also being by used. in and 1984eto reserve statied extensively used in on-farm testing. Kpatcha-Kpatcha byCIMMYT in 1983 and 1984 to research stations in (Poza Rica 7843-SR) and Pirsabak (Pirsabak (1) Soumbe and N'Goulinga; the best varieties in these 7930-SR) from the CIMMYT/IITA joint program trials had yields 60% above the local checks (personal
have been introduced into the southern region. communication, M. Bjarnason). Particularly promising are varieties from Populations 22, 32, and 43.
Maize is the third most important cereal grain,
after millet and sorghum. During 1983-85, approxi- Maize has become a very important staple food for mately 90,000 t of grain were produced annually on most of the population of C6te d'Ivoire. During 1983-85, 128,000 ha, with national yields averaging 0.7 t/ha. approximately 478,000 t of grain were produced annuThe national maize program cooperates with ally on 575,000 ha, with national yields averaging 0.8 t/ CIMMYT, IITA, and SAFGRAD, which is headquar- ha. Both yellow and white flour is used, depending on tered in Burkina Faso. Safita-2, derived from CIM- tribal customs. The grain is used as human food in MYT Pool 16, has been released. In addition, a diverse forms. Recently, more maize is being used as streak-resistant version derived from CIMMYT Popu- animal feed. lation 21 (Tuxpefio-1) is being considered for release in The C6te d'Ivoire national maize program has the southern areas. The Burkina Faso/IRAT maize been developing both improved populations and inbred program selected two liTA hybrids, 8322-13, and 8428-19 lines for a hybrid program (personal communication, in 1985 for commercial release (personal communica- H. Dosso). Improved germplasm from CIMMYT has tion, S.K. Kim) been extensively used since 1975 and 20 experimental
varieties have been developed. Tuxpefio P.B. is widely
CAMEROON grown in the country. The CIMMYT/IITA material
converted to streak virus resistant populations, e.g. EV
Maize is the major food staple in Cameroon. Dur- 8428-SR and EV 8435-SR, has also been used extening 1983-85, approximately 510,000 t of grain were sively. The program requests promising experimental produced annually on 443,000 ha, with national yields varieties from both IARCs to use as varieties per se or averaging 1.2 t/ha. germplasm for their breeding program. The Pioneer
The national maize program cooperates with IITA Hi-Bred International program in Cte d'Ivoire is and CIMMYT. IITA currently has resident maize staff using IITA's streak-resistant lines and has identified a stationed in the country as part of the USAID-supported superior hybrid combination between an IITA inbred National Cereals and Extension Project. The IITA line and an inbred line from Pioneer's maize program staff posted in Cameroon focus their work on varietal in Thailand (personal communication, S.K. Kim). The

designation "FERKE" is used on experimental varie- 2.105-day, white dent varieties for the minor rainy ties, but is changed to "IDSA" when released. season in the forest and transition zone. These meet
In the past, seed production and distribution has about 20% of the national varietal requirement.
been the responsibility of the state seed company, 3.90-day, yellow flint varieties for early planting in BETPA. Recently the Government has provided addi- the Guinea savannah. These meet about 5% of the tional funds to the Ministry of Rural Development to national varietal requirement. improve public-sector seed production and distribu- 4. 90-day soft white dent varieties for use in the tion systems. BETPA has switched from the local land Coastal and Guinea savannas. These meet about 5% of variety CJB to IDSA 29 (Ferke 7928, Pop. 28 Amarillo the national varietal requirement. Dentado) for a late variety and IDSA 28 (Ferke 7635, 5.100-day yellow dent varieties for use throughout Pop. 35, Antigua-Rep. Dominicana) for an early vari- Ghana. These account for about 10% of the national ety (IDESSA 1984). Other varieties that have been varietal requirement. released by the national program and include CIMMYT germplasm are IDSA 26 (Ferke 7529, Pop. 29, To date, three improved varieties, La Posta, ComTuxpefio Caribe), IDSA 27 (Ferke 7622, Pop. 22, posite 4, and Golden Crystal-all full-season white Mezcla Tropical Blanca), and IRAT 81 (a variety dents derived from CIMMYT material-are widely hybrid whose female parent is Tuxpefio brachytic). grown. In addition, an earlier (105-day) white dent, Safita 2 (based on CIMMYT's Pool 16), has been
GAMBIA released for use during the minor rainy season in the
Maize is not a significant crop in Gambia. The area forest and transition zone. Improved open-pollinated planted with maize is 5,000 ha, though the total is varieties are used on a substantial portion of the total growing. Only improved open-pollinated varieties are maize area, and the three above-mentioned full-season planted (personal communication A.G. Carson). The varieties are grown on 25% to 30% of the total national Ministry of Agriculture cooperates with CIMMYT- maize area. Recently, two new varieties based on testing early maturing varieties-and with IITA, testing CIMMYT germplasm have been released: Dobidi CRI early, streak virus, and Striga (witchweed) resistant 1, derived from Ejira (1) 7843, and Aburotia CRI 1, varieties. Research is aimed at improved open-pollinated derived from Tuxpefio P.B. C16. Even though not varieties. Cooperation with the IARCs has been good. officially released, IITA supplied, upon request, 2 t of
The Government's Seed Multiplication Unit d TZESR-W to the Ghana Seed Company in 1983 for
The ovenmet'sSee Muliplcaton nitdis- seed multiplication and distribution in the Volta retributes some seed of improved open-pollinated varie- gion. ties.
Malze is the major food staple in Ghana. During Maize is grown on only about 15,000 ha nationally 1983-85, approximately 402,000 t of grain were pro- (personal communication, Cyril E. Broderick). The duced annually on 418,000 ha, with national yields national maize program receives open-pollinated vaaveraging 1.0 t/ha. Small-scale farmers account for rieties, full-sib progeny, and various disease-resistant 90% of national production. The majority of maize is high yielding hybrids from IITA. It started receiving intercropped, primarily with cassava and cocoyam in variety trials from CIMMYT in 1982, and varieties southern Ghana and with sorghum, millet, groundnuts, from Populations 21, 23, 28, and 43 have performed and cowpeas in the Guinea savannah. well. In addition to testing these materials for local
adaption, some selection is being done locally. The
The Ghana Grains Development Project (GGDP) Central Agricultural Research Institute (CARI) is was launched in 1979 with the objective of making making some crosses. Ghana self-sufficient in maize production (GGDP m osofe reed 1986). As of 1987 CIMMYT had two maize scientists Most of the breeding program is aimed at imand IITA had one cowpea scientist assigned to this proved open-pollinated varieties. Only a very limited project. Objectives of the GGDP maize breeding pro- amount of effort is used in inbred line development. At gram are to develop the following maize materials: present, no releases of improved varieties have been
1. 120-day white dent varieties for the forest and made, though they do have several promising lines. transition zones in the major rainy season. These meet Hybrids are very rare and would make up less than 1% about 60% of the national varietal requirement. of plantings. Maize production has been increasing at

The National Accelerated Food Production Pro gram (NAFPP) in Nigeria conducted a total of 10,21' mini-kit trials in 15 states between 1974 and 1978 NAFPP distributed 728 t of seed of these two varietie. to local farmers (1975 to 1978). Results of the mini-kit trials showed an average 30% increase in yield compared to local farmers' variety. It is estimated that these two varieties are grown on about 1 million ha in Nigeria.
More recently, streak-resistant variety conversion., from CIMMYT Populations 28 (La Maquina 7928) and 43 (Poza Rica 7843) have been released and arc being multiplied by the National Seed Service. Seed multiplication of three additional new varieties has alsc been recently initiated: TZMSR-W for the mid-altitude ecology of Plateau State (combined MSV, H. turcicum and P sorhi resistance) and DMRLSR-W and
DMRESR-W (combined MSV and downy mildew resistance) for the maize areas affected with these diseases in Ondo, Bendel, and Kwara States.
Ajoint IITA-Nigeria program was-as noted previ- ously in the IITA portion of chapter 2-initiated in 1982. In 1984, IITA maize hybrids were tested in on-farm demonstrations (a total of 150 ha) across Nigeria. These plantings were made by Nigerian farmers with the cooperation of both the Nigerian National Figure10. Dr.S.K. Kim, IITA maize breeder, discusses Research and Extension Organizations and the suphybrid breeding work at a government arranged field por h Fedel Ornento a. Selected
day n Lgosstat, Ngera, i 195. ourc: ITA. port of the Federal Government of Nigeria. Selected day in Lagos state, Nigeria, in 1985. Source: IITA. hybrids produced between 25% to 60% more yield
than the check variety, TZSR-W-1. Two private seed arate of2%to4% per year. Seed production is major companies have been recently formed (with IITA asbottleneck to increased production. There is no na- sistance) to produce these hybrids: Agricultural Seeds tional seed production, but a few private companies Nigeria and the Limited and Temperance Seed Comimport seed. pany. Under the supervision of IITA and the National
Seed Service personnel, the two companies produced NIGERIA about 600 t of eight hybrids in 1985 for planting in the
1986 season (personal communication, Y, Efron). In Maize is an important crop in Nigeria. During 1986, approximately 1,600 t of hybrid seed were pro1983-85, approximately 2.1 million t of grain were duced produced annually on 2 million ha, with national yields ducked, enough to plant 90,000 to 100,000 ha. averaging 1.0 t/ha. SENEGAL
Maize improvement under the auspices of IITA
began with the formation and improvement of two Maize is becoming an increasingly important crop populations: TZB, which originated from both African in Senegal. During 1983-85, approximately 101,000 t and Latin American sources, with Nigerian Composite were produced annually on 84,000 ha, with national B most important; and TZPB, which is derived from yields averaging 1.2 t/ha. During the past 5 years, Tuxpefio Planta Baja Cycle 7 from CIMMYT (per- maize production has doubled. The Senegalese govsonal communication, Y. Efron). These two were ernment objective is to attain self-sufficiency (400,000 released by the Nigerian National Program as FARZ t) by 1989. 27 (TZPB) and FARZ 34 (TZB). Both varieties are full The Institut Senegalais De Recherches Agricoles season and have good resistance to tropical rust (Puc- receives material from both CIMMYT and IITA ciniapolysora) and lowland blight (Helminthosporiuin (personal communication, Papa Assau Camara). The maydis). tropical material tested is selected for early maturing

types (90 days, not to exceed 110-120 days). A unique SIERRA LEONE rainy season of 3 to 4 months limits growing of a
full-season crop in Senegal. White flinty grain maize is Maize is not a major crop in Sierra Leone. During preferred for human food, but some yellow and semi-dent 1983-85, approximately 20,000 t of grain were protypes are used by industries. duced annually on 14,000 ha, with national yields averThe local program created a synthetic variety using age 1.4 t/ha.
material from IITA for adaption to African conditions Material for testing is received from both CIMand unrelated genetic material from CIMMYT to MYT and IITA, and cooperation with these lARCs has select for high combining performance with the IITA been good (personal communication, E.R. Rhodes). material. Some CIMMYT materials have been distrib- Three groups of maize trials are received from both uted to farmers after local selection. They include IARCs and these are tested locally. Some work is Camasa 1, based on Tocumen (1) 7835 from CIM- under way to add local adaptation to the most promisMYT, and two quality protein maize varieties based on ing selections. This local research is aimed mostly at Obregon 7740 and Temperate White QPM, also from developing improved local open-pollinated varieties CIMMYT. for farm use. Only 10% of the research is concerned
National germplasm development emphasizes with inbred and hybrid development. Two populations open-pollinated varieties. They have three levels of have been released for farmer use: TZSR (yellow) and selections based on the level of farmer's technology for TZPB (white). All the production is open-pollinated the intended use of the developed variety. One is the varieties. Maize seed is distributed by the Extension creation of varietal hybrids from crosses of Service and private farmers. composites-usually between one local variety and an TOGO introduced material. To obtain the best specific plants,
reciprocal recurrent selection techniques are used. Maize is an important food crop in Togo. During These hybrids are intended for farmers having a high 1983-85, approximately 191,000 t of grain were protechnology for maize production. The second level is duced annually on 179,000 ha, with national yields the creation of synthetic varieties and composites ob- averaging 1.1 t/ha. tained from full-sib selection. These genotypes are The national maize program cooperates with both intended for farmers having an intermediate-level 1ITA and CLMvYT (personal communication, Wilfried technology. The third level is local population im- Schwieber). La Posta, originally from CIMMYT, is the provement by simple recurrent selection methods to most widely grown improved variety in Togo. More develop improved genotypes from local populations recently, two streak-resistant varieties based on CIMfor farmers with low technology. MYT/IITA germplasm, Poza Rica 7843-SR and PirsaThere is a national seed service and a bilateral bak (1) 7930-SR, have been released and are being project with West Germany to promote maize produc- multiplied by a Federal Republic of Germany (GTZ) tion. The research stations furnish foundation seeds to project in Sotouboua. Cooperation with IITA involves the national seed service-German project, and they in improved streak resistance in open-pollinated varieties turn contract with farmers to produce seed. The coop- (both white and yellow grained). CIMMYT openerative project buys the seed and distributes it to the pollinated white and yellow varieties are also being rural extension agencies, which sell directly to their tested. Some research to develop low-input technolofarmers. The maize improvement program recom- gies for the agroforestry zones is also underway. Five
mends the following: open-pollinated varieties have been released to farmers, with four based on CIMMYT germplasm. Ferme
Hybrids: BOS 111, HVB-1 Semenciere de Sotouboua is the national governmental seed company; approximately 120 t of seed is
Varieties: Synthetic C, Camara 1, APM1 and produced per year.
APM2, CP75 white and yellow composites ZAIRE
Local Maize is the major cereal grain produced in Zaire.
population: ZM10 (eastern Senegal) During 1983-85, approximately 709,000 t of grain were
Source: Personal communication, Papa Assau produced annually on 822,000 ha, with national yields Camara. averaging 0.9 t/ha.

The Programme National Mais has cooperated varieties--white and yellow grain-with resistance to closely with CIMMYT and IITA for many years (per- late wilt and stalk rot diseases caused by Cephalosposonal communication, Mulamba Ngandu Nyindu). ium maydis, to H. turcicum leaf blight, and to Ustilago Between 1971 and 1981, CIMMYT had maize staff maydis (personal communication, W. Haag). The naassigned to work with the national program at Lubum- tional maize program has had close collaborative links bashi. It receives the full complement of experimental with CIMMYT for two decades, and several CIMMYT maize staff members have been stationed in Egypt
materials available from the Center. IITA populations during this time period. IITA's inbred lines have been also have been used in crosses with the local Zairian screened by maize researchers in national programs varieties. Populations TZDMRSR-W, TZDMRSR-Y, and private companies for resistance to late wilt (perTZMSR-W, and TZEMSR-W have shown high levels sonal communication, S.K. Kim). of resistance to one or more of the following diseases: The major improved varieties in use are: downy mildew, maize streak virus, and H. turcicuni. 1. Giza-2, an open-pollinated variety with late-wilt Currently, IITA is involved in a substantial bilateral tolerance covering 25% of the total maize area. This maize research assistance project, sponsored by USAID, variety is based on 50% of American Early Dent with the national maize program and has several maize (AED), developed locally, and 50% of tropical germplasm scientists stationed in the country. (Pop. 21) obtained from CIMMYT in the 1969-70
The national program has released four open- period.
pollinated varieties from the CIMMYT material: 2. P-514, a Pioneer hybrid.
3. DC-202, an AED x Tep.-5 hybrid developed by
Salongo: 10 best families of Tuxpefio 1, C11 the national program.
Kasai 1: Tuxpefio x Eto Blanco 4. Kahera-1, a open-pollinated variety developed
PNM 1.: (Tuxpefio and Mix Colima grl I and by Cairo University.
ETO Blanco) x Shaba Safi Certified seed production of these materials during
1984 totaled 8,700 t and was distributed as follows:
DC-202, 2,800 t; Giza-2, 4,400 t; P-514, 1,000 t; and
Kahera-1, 500 t (unpublished report, W. Haag).
Most Zairian farmers plant open-pollinated varie- The national maize program released several new ties. A few imported hybrids are grown in the southern hybrids during 1983-84, all of which are based on AED part of Shaba province. A hybrid program has been x Tep.-5 crosses: DC-204, DC-215 (double crosses); initiated and inbreds are being developed. No seed TWC-9 and TWC-10 (3-way crosses) (unpublished organization exists, and access to seed of improved report, W. Haag). Inbreeding is being conducted in varieties is limited. CIMMYT material in support of national hybrid development efforts. Inbred lines from the following mateNORTH AFRICA AND rials had been developed: Tuxpefio C17; Tlaltfzapan
THE MIDDLE EAST 7844; Gemmeiza 7421; (Ant. xRep. Dom) x Corn Belt;
Sids 7444; Pop. 45; Tuxpefio x ETO; Tep.5; and La
North Africa Posta. Sids 7444 continues to be a source of resistance
to late wilt and to H. turcicum.
Maize is the major cereal grain produced in Egypt. MOROCCO
During 1983-85, approximately 3.6 million t of grain Maize is not a major crop in Morocco. During were produced annually on 805,000 ha, with national 1983-85, approximately 267,000 t of grain were proyields averaging 4.4 t/ha. Virtually all maize is pro- duced annually on 405,000 ha, with national yields duced under irrigation. averaging 0.7 t/ha.
Maize research is carried out by the national maize CIMMYT is cooperating with the Moroccan naprogram, Cairo University, and several private compa- tional maize program to develop high-yielding hybrids nies. National germplasm improvement priorities are and open-pollinated varieties appropriate to dryland as to develop high-yielding hybrids and open-pollinated well as irrigated conditions. Progress has been made in

Maize research began during the 1950s, in collaboration with FAO. Approximately 30 U.S. hybrids were compared with local varieties in multilocational trials. Two double-cross U.S. hybrids, U.S. 13 and Wisconsin 641 AA, showed high yield potential and were released for commercial production. Additional hybrids were developed during the 1950s and 1960s by Turkish maize breeders using local inbreds and lines from the United States.
Because of inadequate production and distribution of hybrid seed, and the subsistence nature of many ,Turkish maize farmers, especially in the Black Sea area, the Turkish national maize program began to give more emphasis to the development of improved open-pollinated varieties (OPVs). Three gene pools were constituted using local and foreign germplasm: S TMP-1, TMP-2, and TMP-3. After making headway in developing OPVs, the Turkish national program renewed its hybrid work in 1980 (personal communicaaN tion, W. Haag). Some 150 public lines from the United States were obtained for evaluation through assistance from the CIMMYT Regional Maize Specialist stationed in Turkey. From these inbred lines, five hybrids
Table 5. Foreign hybrids registered in Turkey in 1984-85
Figure 11. Maize researcher discusses improved Company Name of hybrid
maize with farmer, Morocco. Source: CIMMYT. Northrup King (Tohum Islah) Matador, Mirko, Silco
developing improved materials for the irrigated and Pioneer Inter- P-3360, P-3320, P-3184, higher rainfall areas of the north. U.S. germplasm is national P-3183
playing a major role in the development of hybrids for Lima Grain (Sapek) LG-42, LG-55, LG-66 these areas; Pioneer Hi-Bred is active in this work. Dekalb XL-72AA, DK-789,
Little progress has been made to date toward the DK- 7
development of early, drought-tolerant materials for DK-I-Peron
the low-rainfall areas. Some promising materials for Ciba Geigy G-4524, G-5050
these areas currently under evaluation include BS-5 Basagene (Berket) 61MF, 714MF, 81MF, from the United States and Pools 29 and 30 from Basagene (Bereket) 61MF, 714MF, 81MF, CIMMYT (personal communication, W. Haag). M-7676, M-8161, M-84
Jacques Jx-8820, Jx-247, Jx-187A
Middle East Stauffer S-4460, S-5540, S-6920,
Asgrow RX-90
Maize is less important in Turkey than wheat and Asgrow RX-90
barley, though it is becoming increasingly important as Nickerson Nickerson-702 a poultry feed. During 1983-85, approximately 1.6 million t of grain were produced annually on 560,000 Source: Wayne Haag, unpublished CIMMYT reha, with national yields averaging 2.9 t/ha. port, 1985

have been developed: three single crosses (TTM-813, Table 6. Open-pollinated varieties released by TTM-815, and TTM-815), one three-way cross the Burmese National Maize Program and
(TUM-827), and one double cross (TCM-811). Parent based on CIMMYT germplasm
lines all relate to U.S. public lines derived from Lancaster x Stiff Stalk Synthetic. Grain
Since 1984, many private seed companies with Name color Source material
international operations have established maize seed Shwe-wa 1 Yellow La Calera (1) 7728 production programs in Turkey and have registered a (Pop. 28)
number of "foreign" hybrids for commercial sale (table Shwe-wa 2 Yellow Petrolina 7736 (Pop. 36) 5). Shwe-wa 2 Yellow Petroina 7736 (Pop. 36)
With the growth in the private maize seed sector, Shwe-wa 3 Yellow Satipo (1) 7627 (Pop. 27) the role of the national maize program is being re- Shwe-wa 4 Yellow Indonesian Early viewed. While still under discussion, it appears that Shwe-wa 7 White Taltizapan 7322 (Pop 22) Shwe-wa 7 White Tlaltizapan 7322 (Pop 22)
public sector maize research will concentrate on developing OPVs and improved production practices for the Shwe-wa 8 Yellow Across 7835 (Pop. 35) Black Sea area (personal communication, E. Kinace). Shwe-wa 9 Yellow Pichilingue 7931 (Pop. 31) Important maize improvement objectives include the Shwe-wa 10 Yellow Fareko 8328 (Pop. 28) development of high-yielding hybrids and open-pollinated varieties with resistance to Fusarium stalk rot and H. Shwe-wa 11 Yellow Across 8331 (Pop. 31) turcicum leaf blight and to two stalk borers, Ostringa Source: CIMMYT Maize Program nubilalis and Sesamia cretica.
The major open-pollinated varieties in use are:
Karadeniz Yildizi (K3/74), formed by compositing millet. During 1983-85, approximately 7.8 million t of Yugoslavia hybrids and other germplasm; Ada, formed grain were produced annually, with national yields from CIMMYT Compuesto de Hungaria, Pop. 48; averaging 1.3 t/ha.
Sapanca, formed by compositing Compuesto de Hun- Indian maize research dates back to the colonial gariaand several tropical materials; and Arifye, a late, time and to the work of the Imperial Council of tall variety used mainly for silage. Agriculture. After independence in 1947, maize research continued in various states under the auspices of ASIA the Indian Council of Agricultural Research (ICAR).
A number of inbred lines and double-cross hybrids South Asia were developed in Uttar Pradesh and Punjab states,
notably Punjab-1, Punjab-2, and Punjab-3; these, BuRMA however, were not very high yielding.
In 1957 the Indian Government and the Rockefeller
Maize is the second most important cereal grown in Foundation established a joint Coordinated Maize Burma, and production has been increasing rapidly Improvement Programme for India (Mikoshiba 1971). over the past 15 years. During 1983-85, approximately The Rockefeller Foundation assigned several scien317,000 t of grain were produced annually on 185,000 tists to work with Indian scientists in this new countryha, with national yields averaging 1.7 t/ha. wide maize breeding effort and provided financial
CIMMYT supplies germplasm to the Burmese support and scholarships to strengthen national maize
national maize program (Zin 1986). In particular, the efforts. drought tolerance (partially due to more extensive root One of the first activities of RF and Indian scientists systems) of certain CIMMYT materials has been espe- working in the Coordinated Maize Improvement Procially beneficial. A number of varieties have been gram was to collect and evaluate exotic germplasms gram was to collect and evaluate exotic germplasms
released by the national program, with most based on from other countries in order to broaden the germplasm CIMMYT germplasm (table 6). base available in India. Emphasis was on yellow flint
INDIA grain types. In particular, germplasm complexes from
Latin America proved to be useful, including ETO Maize ranks fifth in importance among the cereals from Colombia, Peru 330, Venezuela 1, Antigua Gr.1 produced in India, after rice, wheat, sorghum, and from Antigua via Mexico, and Cuba 342 (Mikoshiba

Table 7. Background of composite maize varieties released in India in 1967
Name color Germplasm sources
Amber Yellow 11 varietal crosses involving five flint varieties (Narino 330,
Diacol VI, Venezuela 1, Corneli 54, Eto Amarillo) and six dent varieties (Jarvis, Ferguson, Yellow Dent, Cocker 811 Gr/N 1000/1452, Mexican June, Bolita Amarilla)
Vijay Yellow J1 x Coastal Trop. Flints
Jawahar Yellow A1 x Antigua Gr. 1
Kisan Yellow Coastal Tropical Flint
Sona Yellow J1 x Cuba 11J
Vikram Yellow Basi x Eto Amarillo
Source: Maize in India, Haruo Mikoshiba, 1971.
1971). A large number (more than 4,000) of inbred subtropical germplasm for maize varieties and hybrids lines were developed from Indian and foreign adapted to production during the winter (Rabi) season germplasms. By 1961, four double-cross hybrids--Ganga (personal communication, R.L. Paliwal). For example, 1, Ganga 101, Ranjit, and Deccan-had been released a family of Tuxpefio-1 (Pop. 21) is the female parent in for commercial production. the hybrid, Pathari Makka. India has also released an
Although the emphasis in the Coordinated Maize open-pollinated variety, Lakshami, based on CIMImprovement Program was on hybrid development, MYT material (Sids 7444, from Pop. 44) grown extenpopulation improvementworkwas also carried out and sively in Bihar State. A varietal hybrid, Sangam, uses composite open-pollinated varieties and inter-varietal Tuxpefio Planta Baja (C7) as one parent. Finally, hybrids were developed. Outstanding among these J115-"Parbhat," is based on Suwan 1 from Thailand. populations were J-1 (also known as Naraingarh Since 1985, IITA has also supplied its inbred lines and Complex), and Sona (J1 x Cuba 11J). In 1967, six hybrids for testing in India. high-yielding composites--Amber, Vijay, Jawahar, Kisan,
Sona, and Vikram-developed from these populations NEPAL were released for commercial production (table 7). Maize is the second most important cereal proIn addition to national breeding stocks, Indian duced in Nepal after rice. During 1983-85, approximaize researchers have made considerable use of mately 784,000 t of grain were produced annually, with germplasm from Colombia (ETO germplasm com- national yields averaging 1.5 t/ha.
plexes) and southeastern United States (North Caro- The Nepalese National Maize Development Prolina and Florida) in developing inbred lines (IARI gram (NMDP) receives a broad range of germplasm 1980).While dozens of hybrids have been released, the from CIMMYT (Sharma and Anderson 1985). Most area planted with hybrids remained stagnant at about farmers grow open-pollinated varieties, though a few 10% to 15% through the 1970s and early 1980s. This farmers in the south grow Indian hybrids. Three situation has begun to change with the expansion of open-pollinated varieties have been released from privatemaizeseedsectoractivities.Asof 1986,some10 CIMMYT materials after local selection for better private seed companies (transnational and national) adaptation to Nepalese farming: are engaged in maize improvement research and seed 1. Janaki (white): (Rampur 7434; Pop. 34) production in India (Pray 1986). These private sector 2. Arun-2 (yellow): (UNCAC x Phil DMR; Pop. 59) hybrids are grown on about 10% of the total maize
area. 3. Makalu-2 (yellow): (Amarillo del Bajio; Pop.45)
Indian maize researchers participate in CIMMYT's Of these OPVs, Janaki and Arun-2 have been the International Maize Testing Program. Perhaps CIM- most popular with farmers (Sharma and Anderson MYT's biggest contribution has been in providing 1985).

Table 8. Open-pollinated varieties released in Pakistan, 1970-85
Variety color Germplasm sources
Changez White Swabi white (local) and U.S. inbred lines (WF9, M14, B37,
WM13R, HY, A619, Oh45)
Zia White (Early King x Payette) x Changez
Khyber White Akbar x Bowman's Cole Creek
Shaheen White Zia x (Nodak, Mandan, Payette)
Sadaf White White version of Neelum
Sarhad White White Akbar x Tropical x U.S. Corn Belt
Azam White Zia x Pirsabak 7930
Ehsan White Sarhad White x Pirsabak 7934
Neelum Yellow Local x Latin American x U.S. Corn Belt materials
Agaiti-72 Yellow Six-way cross (M14 x Pa32) x (WF9 x W9) x (A495 x A556)
Akbar Yellow Neelum x U.S. Corn Belt materials
Sharhad Yellow Yellow Yellow version of Sarhad
Sultan Yellow Akbar x (Syn 548, Syn 547, Neluum and Comp. II)
Sunheri Yellow Agaiti-72 x Amarillo Cristalino-2
Kashmir Gold Yellow Yousafwala 7845 (Pop.45, Amarillo Bajio)
Source: National Coordinated Maize Programme, Pakistan Agricultural Research Council.
PAKISTAN oped, based primarily on U.S. temperate germplasm
Maize is the third most important cereal produced and some tropical and subtropical materials from in Pakistan, after wheat and rice. During 1983-85, CIMMYT (table 8). approximately 1.0 million t of grain were produced on CIMMYT and the National Coordinated Maize 801,000 ha, with national yields averaging 1.3 t/ha. Program of the Pakistan Agricultural Research CounMost of the maize area is irrigated, and the crop is cil have had close collaborative research ties for two grown in the hot summer season under very difficult decades. CIMMYT resident maize advisers have been climatic conditions. stationed in Pakistan since 1968, and considerable
Initial maize improvement work in Pakistan fo- amounts of CIMMYT germplasm are evaluated annucused on hybrid development, relying primarily on ally. Since 1985, national maize researchers have also inbred lines and hybrids from the U.S. Corn Belt received IITA's white and yellow grain hybrid trials. supplied through USAID during the 1950s. In the mid- It is estimated that 30% to 40% of the farmers use 1960s, the hybrid development work was discontinued some kind of mixtures of local varieties and improved due to the lack of a viable seed industry to produce and oe o te o etie n re distribute hybrid seed. Emphasis was placed on devel- genotypes. Among the most extensively grown are Zia,
disribte ybrd sed.Emhass ws pace ondevl- Shaheen, Sarhad, Azam, Akbar, Neelum, and Sultan oping open-pollinated varieties (synthetics, compos- Shaheen, Sarhad, Azam, Akbar, Neelum, and Sultan ites), largely based on temperate germplasm but with (Chaudiry 1984). Three open-pollinated varieties based some tropical germplasm introgressed for disease on CIMMYT materials have been released in Pakiresistance to leaf blights (H. turcicum and H. maydis) stan: Azam, Ehsan, and Kashmir Gold (not officially and stalk rots caused by Fusarium spp. A number of released but widely grownin the States ofAzad Jammu outstanding open-pollinated varieties have been devel- and Kashmir).

Figure 12. Maize trial conducted at the Maize and Millets Research Institute, Punjab, Pakistan. Source: CIMMYT.
Provincial and federal maize research programs distributes its hybrids to contract growers who produce have reactivated their hybrid development programs in grain for wet-milling. recent years (PARC/CIMMYT 1988). The primary
germplasm sources used in these hybrid programs are: SRI IANKA U.S. public inbred lines such as Mo 17, B-73, B-44,
A-619, Oh 43, Oh 45, and H-51; older local inbred lines Maize is the second most important cereal crop such as Punjab-7, and new inbred lines extracted from (after rice) in Sri Lanka. It is produced on about 50,000 A Iha and is used mainly for animal feed. Over 90% of the national breeding populations and commercial vae- area is in the highlands and is irrigated. Maize breeding
ties. objectives include resistance to stem borer (Chilo
The lack of a viable seed industry remains a major partellus), stalk rots (Fusarium sp.), leaf blights (H. constraint in the delivery of improved genotypes to turcicum), and sheath spots. Populations are being farmers. Only enough certified seed to plant 5% of the formed using local materials and germplasm provided total nationalmaize area is produced annually. Most of by CIMMYT (personal communication, R.N. Wedthe certified seed is of open-pollinated varieties. Only derburn). One of the most promising populations, about 13,000 ha are planted with hybrid maize in Composite 6, is comprised of Thai composite, CupuPakistan t of this area is planted with hybrids rico x flint compuesto, and Poza Rica 7425. Various Pakistan. Most of this area is planted with hybrids varieties from CIMMYT's populations 28, 29, 36, and
developed by Rafhan Maize Products, Ltd., a corn 41 have also been introduced. Open-pollinated variestarch company. Rafhan's hybrids are based on U.S. ties and nonconventional hybrids (varietal and family inbred lines (MO17, B73) as well as locally developed crosses) are being developed. A yellow-grain variety inbreds with resistance to leaf blights and stalk rots developed from Suwan 1 was released in 1977 under (personal communication, Khan Bajadur). Rafhan only the name of Bhadra 1.

Southeast Asia and Pacific Baflos, has participated in CIMMYT's international
variety testing program since the mid-1970s (personal
INDONESIA communication, M. Latin). IPB's maize breeding program has been primarily geared toward the developMaize is the second most important cereal pro- ment of superior open-pollinated varieties (composduced in Indonesia, after rice. During 1983-85, ap- ites). A modest hybrid breeding component of the proximately 5.2 million t of grain were produced annu- program was initiated in 1980, and inbred lines have ally on 2.6 million ha, with national yields averaging 2.0 been extracted from several source populations. Some t/ha. Average maize yields have been increasing in hybrids are being used in limited areas covered by the Indonesia due to the rapid increase in the use of government's intensified maize production program. fertilizer, especially nitrogen, which is being sold at Although CIMMYT has been a major source of highly subsidized prices. germplasm, most of its material has lacked resistance
CIMMYT collaborates with the national maize to downy mildew, a major constraint to maize producprogram and regularly supplies international testing tion in the country. IPB has also participated in a trials and makes staff visits to the country. More CIMMYT-coordinated project to improve several recently, IITA has supplied inbred lines with combined populations for downy mildew resistance, using Philipdowny mildew and streak virus resistance to the na- pine and Thai source materials. They have tested tional program and to a private seed company, P.T. experimental varieties with tropical adaptation. In 1986, Bright. IPB released its first variety based on CIMMYT
The maize seed industry is not well developed and germplasm, IPB Variety 4, a white flint based on only about 25% of the total maize area is planted with Rampur 8075, derived from a downy mildew resistant improved varieties. In the future, it is envisioned that population developed specially through the collaboraboth open-pollinated varieties and hybrids will have tive IPB-CIMMYT program. Since 1985, IITA has market niches. Two government agencies-Perem Sang supplied inbred lines and populations to IPB and two Hyang Seri and P.T. Pertaini Patra Tani-produce private companies, the San Miguel Corp. and Pioneer some certified maize seed of open-pollinated varieties. Hi-Bred International (personal communication, S.K. In recent years, a number of private seed companies Kim). have also been established. Three transnational Currentlyfour seed companies (one local and three companies-P.T. Bright (division of Dekalb), Cargill, foreign) produce hybrid maize seed in the Philippines. and Pioneer Hi-Bred-have released downy mildew These companies produce hybrids that undergo an resistant hybrids. Cargill's hybrid, C-1, is the most accreditation process based on agronomic performpopular, and the seed demand for this genotype has ance trials coordinated by the Philippines Seed Board. outpaced supply (Timmer 1987). P.T. Bright is also Pioneer Hi-Bred has two downy mildew tolerant hyselling an open-pollinated variety, Arjuna, which is brids that are officially approved. Pioneer Hybrid 6181 derived from Suwan 2 developed at Kasertsart Univer- is a yellow material released in 1980 and extensively sity in Thailand. grown in Southeast Asia. Pioneer Hybrid 3228 was
PHILIPPINES released in 1985; it is better yielding than 6181 and has
better tolerance to downy mildew. Parent lines in 6181
Maize is the second most important cereal pro- were derived from Thai Composite #1, Thai Comp. duced in the Philippines, after rice. During 1983-85, 3-4F4, and Cupurico x Flint Composite DMR. Cargill approximately 3.3 million t of grain were produced also has a downy mildew resistant hybrid released in annually on 3.4 million ha, with national yields averag- 1983. The San Miguel Corporation has released two ing 1.0 t/ha. Most improved cultivars (yellow and varietal hybrids with downy mildew resistance: SMC-101 white) grown by farmers are open pollinated. Of the (EV from Pop. 28 x Suwan 1) and SMC-102 (EV from approximately 3.4 million ha annually planted with Pop. 36 x Suwan 1). maize, only 1% of the total maize area is planted with Certified seed may be marketed either directly by hybrids and about 25% is planted with improved the company or through local marketing companies open-pollinated varieties, and agricultural input suppliers. The open-pollinated
Maize improvement research is conducted in the varieties are produced and marketed by a limited publicand privatesectors. The Institute ofPlantBreed- number of small seed growers. These varieties do not ing (IPB) at the University of the Philippines, Los benefit from a good seed production and distribution

Figure 13. Drs. Rachain Thiraporn, manager of Suwan Farm of Kasetsart University in Thailand, and Richard N. Wedderburn, a CIMMYT entomologist stationed in Thailand, view a test plot of corn. Much of the research on corn improvement in Thailand has been carried out at Suwan Farm. Source: CIMMYT.
system, and hence the rate of spread has been slow. and private sector organizations in Thailand. Although CIMMYT and IITA materials have been used in formTHAILAND ing new populations and in extracting inbred lines, no
Maize is the second most important cereal grain direct releases have been made from these materials. produced in Thailand, and the country is a net maize The history of the development of Suwan 1 is th( exporter of more than 2 million t annually. During epitome of what international cooperation in germplasnr 1983-85, approximately 4.2 million t of grain were development can achieve and has achieved. The stor3 produced annually on 1.7 million ha, with national starts with the introduction in 1960 of the variety yields averaging 2.4 t/ha. Tiquisate Golden Yellow (a mixture of two strains ol
The Thailand National Maize Program has coop- maize of Cuban origin-a white semident and a golden erated with CIMMYT since the mid-1960s and with yellow flint) from Guatemala. This variety was well various Rockefeller Foundation-supported collabora- adapted to Thailand and had a golden yellow semi-flint tive maize research programs in Asia and in other grain that was well regarded in international markets. continents. The CIMMYT international trials have Farmers took up this variety, and the maize area in been tested extensively in Thailand but have lacked Thailand began to increase rapidly. suitable resistance to downy mildew. More recently, Based on the adaptation of Tiquisate Golden YelIITA has supplied its inbred lines and hybrids to public low in Thailand, the national program requested anc

Table 9. Germplasm assembled in Thai Composite #1
Source Group Material
Caribbean Islands Argentino Cuba Gr.l
Argentino Cuba 11J
Argentino Puerto Rico Gr.l
Tuson Cuba 40
Tuson-Carilla-Criollo-Tuson Cuba IJ
Tuson-Carilla-Criollo-Tuson Cuba V59
Tuson-Carilla-Criollo-Tuson Antigua Gr.l
Tuson-Carilla-Criollo-Tuson Antigua Gr.2
Tuson-Carilla-Criollo-Tuson Puerto Rico Gr.2
Tuson-Carilla-Criollo-Tuson Barbados Gr.l
Tuson-Carilla-Criollo-Tuson Cupurico
Tuson-Carilla-Criollo-Tuson Caribb. Flint Comp.
Tuson-Carilla-Criollo-Tuson Comp. Caribb. Amarillo
Tuson-Carilla-Criollo-Tuson Tiquisate Golden Yellow x Caribb.
Tuson-Carilla-Criollo-Tuson Tiqufsate Golden Yellow x Guadalupe
Meidco and Tuxpefio Veracruz 163
Central America Tuxpefio Veracruz 181
Tuxpefio Veracruz Gr.48
T Xpefio Tamaulipas 8
Salvadoreflo Salvadorefio Amarillo
Argentino-Criollo Tiqufsate Golden Yellow
South America Northern Catato Guayana Francesca III
Cuban Yellow Dent Bahia III BCO
Cuban Yellow Dent Dentado Amarillo
Argentino-Criollo-Tuson Narifio 330-Peru 330
Argentino-Criollo-Tuson DV 103
India Caribbean-Tuxpeflo-India-U.SA. Composite Al
Caribbean-Tuxpeflo-India-U.SA. Multiple Cross 2
Caribbean-Tuxpefio-India-U.SA. Multiple Cross 4
Caribbean-Tuxpeflo-India-U.SA. Synthetic A3B
Caribbean-Tuxpeflo-India-U.SA. Synthetic All
Other Tuxpefio-Caribbean-U.SA. Tuxpantigua
Tuxpefio-Caribbean-U.S.A. Veracruz 181 x Antigua Gr.2
Tuxpefio-Caribbean-U.S.A. USantigua.
Tuxpefio-Caribbean-U.S.A. Florida Synthetic
Source: Sujin Jinahyon (1973)

received other tropical varieties and collections of high-yielding, downy mildew resistant material has Caribbean origin from the Mexican-Rockefeller Foun- been crossed with elite inbred lines from the United dation program in Mexico. This germplasm develop- States, Brazil, the Philippines, and elsewhere to proment process was accelerated in 1966 with the shifting duce excellent hybrids and varieties for tropical envifrom India to Thailand of the Rockefeller ronments. Cargill released an open-pollinated variety
Foundation-supported Inter-Asian Corn Program and in 1985, Cargill 357, based on Suwan 1 and South the establishment of CIMMYT; both programs played American germplasm. Pioneer Hi-Bred has released an active role in the formulation of new national the Hybrid 3228, developed in the Philippines. Kasetbreeding materials. sart University has also released a single-cross hybrid
In 1967, the Thai national program chose 36 and a three-way cross hybrid based on the Suwan 1 germplasm complexes to form Thai Composite #1. germplasm complex. General yield performance, adaptation in Thailand, Improved maize seed is produced by the Departand diversity of material by origin were the major ment of Agricultural Extension, Ministry of Agriculcriteria used in the selection of these germplasms ture and Cooperatives, and by several private seed (table 9). companies. During 1984-85, it was expected that beA modified ear-to-row scheme was employed to tween 10,000 t and 13,000 t of certified maize seed of systematically inter-mate all germplasm sources. In open-pollinated maize varieties would be produced. 1970, four cycles of selection were completed, and the Hybrid seed production has been increasing very rapcomposite was then officially named Thai Composite idly. In 1985, about 2,500 t was produced, compared to #1. By 1973, using S, progeny selection followed by 30 t in 1981 (Pray 1987). recombination, three cycles of selection had been
completed. In addition, 250 S1 families were chosen VIETNAM during cycle 3 selection to form the sub-population Maize is the second most important cereal proThai Composite #1 Early, which eventually became duced in Vietnam. During 1983-85, approximately 500,000 Suwan 2. t of grain were produced annually on 387,000 ha, with
While Thai Composite #1 was 20% to 30% higher national yields averaging 1.3 t/ha.
yielding than Tiquisate Golden Yellow, it faced a new CIMMYT supplies germplasm to the Vietnamese problem by the time it was theoretically ready for national maize program and two open-pollinated varelease. Downy mildew, first reported in 1968, had rieties based on these materials have been released: spread rapidly across the maize growing areas of Thai- VM-1 (C7 of Tuxpefio-1, Pop. 21) and MSB-49 (Pop. land. Thai Composite #1, as were other Thai materi- 49). Also, Suwan-2 has been released as TSB-2 and the als, was susceptible to downy mildew. Ganga 5 Composite as Ganga 5. In 1985, it was estiStopgap measures were taken in 1973, and two mated that 80,000 hain northern Vietnam were planted varieties with some resistance to downy mildew, Bogor with VM-1 (personal communication, B.L. Renfro). A Synthetic 2 and Tainan 10, were imported for use by quality protein maize variety, Population 63, has been Thai farmers. Simultaneously, several sources of downy released (NRC 1988). MSB-31 (originally Suwan mildew resistance found in Philippine germplasm, DMR 8331) is under increase and has been approved for
1 and DMR 5, were backcrossed three times to the release. Thai Composite #1 in 1972 and 1973 (personal communication, B.L. Renfro). East Asia
The new material, called Thai Composite #1 DMR, CHINA combined high yield potential with the needed resistance to downy mildew; this was named Suwan 1 and Maize is the third most important cereal grain in released in Thailand in 1974. Since then, Suwan 1 has China, after rice and wheat, and China is the leading spread quickly to farmers' fields in Thailand, where it maize producer in the Third World. During 1983-85, is still the major variety grown by farmers. It has also approximately 68 million t of grain were produced been released in some 20 countries of Asia, Latin annually on 18.4 million ha, with national yields averagAmerica, and Africa (personal communication, B.L. ing 3.7 t/ha. Renfro). Nearly two-thirds of China's maize area is temperSuwan 1 has also been used extensively by private ate in environment with the remaining area plant breeding companies in hybrid development. This subtropical-to-tropical in its growing environment. The

main maize growing areas are the north central region LATIN AMERICA with 8.4 million ha, the northeastern region with 5.6
million ha, and the southwestern region with 4.5 mil- Mexico, Central America, lion ha. Most of the crop is planted on less soils in and Caribbean
northern China and under rainfed production conditions. The northeastern region has a climate similar to COSTA RICA the U.S. Corn Belt.
Hybrids based on U.S. dent germplasm are used by Maize is the second most important cereal proHybrid bo7 e on Tduced in Costa Rica, after rice. During 1983-85, ap65% to 70% of farmers. The original double-cross proximately 105,000 t of grain were produced annually hybrids introduced during the 1950s and 1960s were on 62,000 ha, with national yields averaging 1.7 t/ha. based on elite Chinese Flint germplasm crossed with Maize research is conducted by the University of U.S. Corn Belt lines, including 38-11, L289, W19, W 20, Costa Rica and the Ministry of Agriculture; both orW24, and M14 (Li Jingxiong 1987), By 1969 national organizations cooperate with CIMMYT in germplasm maize breeders had stopped using older tines of both development (Stewart 1985). Nationally released maforeign and local origin. New inbreds have been devel- terials based on CIMMYT germplasm include TICO oped with multiple disease resistance. Most hybrids in V-1 Mejorado (Tuxpeflo P.B., C11, Pop. 21), TICO V-2 China are single crosses derived from these newer (Mezlca Amarilla, Pop. 26), TICO V-5 (Pop. 26), and inbred lines. Of these genotypes, the widely adapted, Diamantes 8043 (La Posta). In addition, avariety-cross the high-yielding hybrid Zhongdan 2, has been the hybrid based on Pop. 26 x Pop. 21 has been proposed most popular, covering up to 9.3 million ha at one time for national release. In 1985,90 t of certified seed were or other during the 1980s; this hybrid has good resis- produced. tance to leaf blights and head smut.
Since 1974, Chinese maize researchers have also CUBA
been breeding for improved protein quality in maize Maize is the third most important food crop in and expect high-lysine and high-oil maize will be util- Cuba, after rice and sweet potatoes. During 1983-85, ized for livestock feeding and industrial purposes. A approximately 97,000 t of grain were produced annuquality-protein maize variety, Tuxpefio 102, has been ally on 67,000 ha, with national yields averaging 1.3 t/ released (NRC 1988). Approximately 10 hybrids have ha. been developed and have undergone regional testing. Maize research in Cuba began more than 40 years The best QPM hybrid available to date, Zhongdan 206, ago and during the 1940s and 1950s produced some yieldedwithin 4% of the normal protein hybrid, Zhongdan well-known and widely used genotypes such as Cuba 2, and also showed good resistance to kernel rot (Li Yellow and Poey T-66. In the 1960s, little maize reJingxiong 1987). search was undertaken. During the 1980s, maize reCIMMYT has had increasing collaboration with search and production has received a higher priority in Chinese maize research programs over the past 10 Cuba's 5-year development plans. As such, CIMMYT yeas.Chinee a relesec twog o verate vaieties collaboration is relatively new, beginning in 1982 (Sanchez years. China has released two open-pollinated varieties and Scobie 1986). CIMMYT regularly supplies based on CIMMYT's Tuxpefio materials: Mexican germplasm to the Cuban national maize program, white-94 (Tuxpefio-1) and Mexican white-1 (Tuxpefio which has released three open-pollinated varieties based P.B. C15). It is estimated that 100,000 ha-5% of on this material: Pichilingue 7928 (Pop. 28), Across China's more tropical maize-growing area-are planted 7926 (Pop. 26), and Across 7931 (Pop. 31). with Mexican white-94 and Mexican white-1 (Li Jingxiong
1987). China has also used CIMMYT's Population 28 EL SALVADOR (Amarillo Dentado-2) to develop one of the inbred Maize is the major food staple in El Salvador. lines in Guangxi Top 4, a hybrid released in 1985. In During 1983-85, approximately 497,000 t were proaddition, the male parent of Hybrid 203 is an inbred duced annually on 242,000 ha, with national yields line (#079) extracted from CIMMYT quality protein averaging 2.1 t/ha, the highest in the region. White-grain maize Population 39. These materials are being grown hybrids are the dominant maize materials grown in the in the southern part of the countrywhere more tropical country. environments prevail. In 1986, the Chinese Academy The Agricultural Technology Center (CENTA) is of Agricultural Sciences also received 100 inbred lines responsible for maize improvement research in El from IITA as germplasm sources. Salvador. It has had a long and close working relation42

Table 10. Tropical lowland maize varieties and hybrids released in Guatemala and containing CIMMYT germplasm, 1972-86
Variety Grain
name Type color Source material
ICTA B-1 OP White Tuxpefio-1 (Pop. 21)
ICTA B-3 OP White Tuxpefio-1 (Pop. 21)
ICTA B-5 OP White Blanco Cristal.-2 (Pop. 30)
La Maquina OP White Mezcla Trop. Blanca (Pop. 22)
ICTA T-101 VC White Tuxpefio-1 x Eto Blanco (Pop. 32)
HB-10 FC White Male parent, Pop. 22
HB-11 FC White Female parent, Pop. 22
HB-19 FC White ICTA B-1 x Eto. Blanco
HB-33 FC White Family of Pop. 23
HB-67 FC White Family of Pop. 22
HB-69 FC White Family of Pop. 23
HB-83 DC White Family of Pop. 22 and inbred lines of Pop 29
ICTA A-4 OP Yellow Am. Cristal.-2, (Pop. 31)
ICTA A-6 OP Yellow Am. Dent. (Pop. 28)
HA-28 FC Yellow Family of Pop. 28
HA-44 VC Yellow Pool 21 x Ant. Ver. 181
Nutricta OP 02 White QPM Tuxpefio
Source: National Maize Program, Institute of Agricultural Science and Technology, Guatemala.
OP = open-pollinated variety; VC = variety-cross hybrid; FC = family-cross hybrid; DC = double-cross hybrid; OPV 02 = quality protein maize open-pollinated variety
ship with CIMMYT. Work to develop hybrids began in approximately 1.1 million t were produced annually on the 1960s. Today, three CENTA-developed hybrids, 777,000 ha, with national yields averaging 1.4 t/ha. H-3, H-5, and H-8, are grown on 71% of the total Maize is mostly grown by small land holders for direct national maize area. The female parent of these hy- food consumption. Guatemala is currently self-sufficient brids is derived from Tuxpefio material. The most in maize and may soon become an exporter. recently developed hybrids are H-19, H-20, and H-33, Corn research is conducted by the Institute of which are now being tested by farmers. H-19 yields Agricultural Science and Technology (ICTA). ICTA 30% more than H-3 and 10% more than H-5 and has was created in 1973 and organized outside the Ministry ear rot resistance. New corn stunt resistant hybrids are of Agriculture. One of its main objectives was to bring in the experimental stage and appear very promising. research into closer contact with both farmers and
El Salvador is one of the few developing countries extension agents. Both USAID and the Rockefeller that is not only self-sufficient in maize seed but also Foundation were involved. CIMMYT placed two sciexports seed to neighboring countries. In 1985,3,000 t entists with ICTA for several years to help strengthen of certified seed were produced. This success is due to the maize research program. These scientists left a a well-focused national maize research program, a working base that ICTA's scientists have built on. dynamic and progressive private seed industry, and a strict and efficient certification system. ICTA has developed a considerable number of
high-yielding varieties and hybrids for lowland tropical GUATEMALA areas based on CIMMYT materials (table 10). A
Maize is the major food staple in Guatemala, and quality protein maize variety, Nutricta, was also rewhite grain is the preferred color. During 1983-85, leased in 1984. In 1985, 2,000 t of certified seed were

Figure 14. Entries in maize trials in Haiti. Source: A.I.D.
produced, with almost all of this used in the tropical have been selected for national release-Carolina, based lowland areas. It is estimated that 60% (120,000 ha) of on Poza Rica 7427 (Pop. 27); La Maquina 7827 (Pop. the maize area in the western lowlands of Guatemala is 27); and La Maquina 7928 (Pop. 28). In 1985, 50 t of planted with improved varieties and hybrids, certified maize seed were produced.
For the highlands, five improved open-pollinated
varieties also have been developed using local materi- HONDURAS als: V-301 (white), Barcena 71 (yellow), V-302 and Maize is the most important cereal produced in V-304 (yellows), and Chanin, which has a cycle of 160 Honduras. During 1983-85, approximately 520,000 t of days compared to 270 days for local materials (Stewart 1985). To date, no significant coverage has been achieved grain were produced annually on 353,000 ha, with with these highland materials, national yields averaging 1.5 t/ha. Currently, the country is a net exporter of maize.
HAIM The national maize improvement program has a
long history of cooperation with CIMMYT, and previMaize is the most important cereal produced in ously with the Rockefeller Foundation. A number of Haiti. During 1983-85, approximately 147,000 t of grain open-pollinated varieties have been developed using were produced annually on 157,000 ha, with national germplasm supplied through this international coopyields averaging 0.9 t/ha. eration (table 11).
CIMMYT has had research staff in residence in the
country since 1982, working with national maize program staff. CIMMYT supplies germplasm to the na- Maize is the major food staple in Mexico. During tional program, and three open-pollinated varieties 1983-85, approximately 13.8 million t of grain were

Table 11. Maize varieties released in Honduras and containing CIMMYT germplasm, 1958-85
Variety Grain
name Type color Source material
Sintetico Hondurefio Tuxpefio OP White Tuxpefio, several populations
Honduras B-104 OP White Tuxpeiio-1 (Pop. 21)
Hondurefio Planta Baja OP White C17 of Tuxpeflo P.B.
Guaymas B-101 OP White La Maquina 7422
Guaymas B-102 OP White Tlaltizapan (Pop. 43)
Guayape B-103 OP White Blanco Cristalino-2 (Pop. 30)
Guaymas A-501 OP Yellow Tocumen 7428
Guaymas A-502 OP Yellow Amarillo Cristalino-2 (Pop. 31)
Source: CIMMYT Maize Program
produced annually on 8.2 million ha, with national a considerable number of improved open-pollinated yields averaging 1.7 t/ha. varieties and variety-cross hybrids have been released
Although CIMMYT is headquartered in Mexico by INIFAP for lowland, tropical conditions (table 12). and shares historical ties with the national research CIMMYT estimates that about 2.5 million ha are system (Stakman et al. 1967), maize research collabo- planted with these varieties; V-524 and VS-525 are the ration between CIMMYT and national institutions has most widely grown. been quite weak until recently. Since the mid-1980s, In recent years, more emphasis has been placed on CIMMYT has strengthened its collaborative ties with germplasm development for highland areas above 1,500 the maize improvement and crop management re- m. The CIMMYT-German Pool STR is providing search programs of the National Forestry, Agriculture, useful germplasm to develop materials for the high and Livestock Research Institute (INIFAP). The na- elevations of the State Chihuahua. CIMMYT is suptional maize program tests CIMMYT's full range of porting Mexico's highland germplasm development germplasm undergoing improvement and conducts work through increased maize improvement activity yield trials throughout Mexico. Out of these materials, for zones above 1,800 m.
Table 12. Maize varieties and hybrids released in Mexico and containing CIMMYT germplasm, 1970-85 Grain
Name Type color Source material
V-424 OP White Blanco Dentado-2 (Pop. 49)
V-425 OP White Pool 19
V-454 OP White Tuxpefio-Caribe (Pop. 29)
V-455 OP White Mezcla Trop. blanca (Pop. 22)
V-524 OP White Tuxpefio-1 P.B. (Pop. 21 C11)
VS-525 OP White La Posta (Pop. 43)
V-526 OP White Poza Rica 7843 (Pop. 43)
V-527 OP White Across 7926 (Pop. 26)
V-528 OP White Tropical Blanco Dentado
HV-313 VC White Blanco Dentado-2 x Lucio Blanco
Source: CIMMYT Maize Program.

Table 13. Open-pollinated varieties released in Maize breeding is carried out by IDIAP, the naNicaragua and containing CIMMYT tional agricultural research institute, and the Univergermplasm, 1975-85 sity of Panama. Close collaboration between CIMMYT and these institutions has existed for many years
Grain (personal communication, Jos6 C. Cedefio). Materials
Name color Source material received from CIMMYT go through a recurrent selecNB-100 White Jutipa 7930 (Pop. 30) tion program and testing for adaption in different parts
of the country. Two open-pollinated varieties from this
NB-1 White Tuxpefio-1 (Pop. 21) program have been released and are planted on a very
NB-3 Yellow La Maquina 7422 (Pop. 22) small acreage. These are Tocumen 7428 and an interNB-4 Yellow Poza Rica 7822 (Pop. 22) varietal hybrid FAUP-1, both from CIMMYT-based
material. Pioneer Hi-Bred International is the only
NB-5 White La Maquina 7843 (Pop. 43) company from which the government buys maize seed
NB-6 White Santa Rosa 8043 (Pop. 43) for resale to commercial users. Two Pioneer hybrids
are used: X-304C and X-306B. During 1985, 93 t of
Source: CIMMYT Maize Program certified maize seed were sold, mostly of these imported hybrids.
In addition, the national maize program is collabo- Andean Region, South America rating with CIMMYT's recently initiated hybrid program. Approximately 6,000 lines have been provided BoLIvIA for testing as potential parents for hybrid development. Maize is the major cereal produced in Bolivia. Various private seed companies are also active in During 1983-85, approximately 460,000 t were proMexico: Pioneer Hi-Bred, Dekalb, Asgrow, Northrup duced annually on 311,000 ha, with national yields King, and Cargill. These companies use CIMMYT averaging 1.5 t/ha. germplasm extensively in their maize improvement CIMMYT cooperates with the national maize programs. It is estimated that 800,000 ha are planted improvement program, and three open-pollinated with hybrids sold by these companies (personal com- varieties have been released that are based on CIMmunication, Alejandro Ortega C.). MYT material: CIAT V105, based on Poza Rica 7528;
Tuxpeflo 02, based on Tuxpefio 02 (Pop. 37); and
NICARAGUA Chuquisaca 7741, based on OPM Population 41.
Maize is the major food staple in Nicaragua. During 1983-85, approximately 213,000 t of grain were COLOMBIA produced annually on 172,000 ha, with national yields Maize is the second most important cereal proaveraging 1.2 t/ha. duced in Colombia (after rice). During 1983-85, apThe national maize improvement program cooper- proximately 870,000 t of grain were produced annually ates with the CIMMYT testing program. Five on 592,000 ha, with national yields averaging 1.5 t/ha.
open-pollinated varieties have been released for com- The Rockefeller Foundation and Colombian Minmercial production after three to four testing cycles istry of Agriculture started an agricultural program in and some selection for reduced height and improved the early 1950s and a strong maize breeding program husk cover (table 13). was initiated (Stakman et al. 1967). Several outstandIn 1985, 1,636 t of certified maize seed were pro- ing germplasm complexes were assembled during the duced; open-pollinated varieties make up 85% and 1950s (see chapter 2), and a number of high-yielding hybrids 15% of this total (personal communication, varieties and hybrids were released for commercial H.H. Lizarroga). Two imported hybrids, DeKalb B-666 production. and Pioneer 5065-A, are also used. The maize program of the Instituto Colombiano
Agropecuario (ICA) collaborates with the internaPANAMA tional germplasm development and testing network
Maize is the second most important cereal pro- coordinated by CIMMYT. CIMMYT's Population 22 duced in Panama, after rice. In 1984, 80,000 t of grain (Mezcla Tropical Blanca), Population 26 (Mezcla were produced on 75,000 ha annually, with national Amarilla), Population 27 (Amarillo Cristalino), Popuyields averaging 1.1 t/ha. lation 35 (Antigua x Republica Dominicana), Popula46

Table 14. Maize varieties and hybrids released in Colombia and containing CIMMYT germplasm
Variety Type color Germplasm sources
ICA V-109 OP Yellow Suwan 7726 (Pop. 26), La Maquina 7827 (Pop. 27),
Tocumen (1) 7735 (Pop. 35), and Santa Cruz Porillo (1) 7835 (Pop. 35)
ICA V-155 OP White Tuxpefio 1 (Pop. 21)
ICA V-156 OP White Tuxpefio 1, C14 (Pop. 21)
ICA V-157 OP White Across 7522 (Pop. 22)
ICA V-258 OP White Tuxpefio Braquiticos
ICA H-353 VC White Tuxpefio P.B. (Pop. 21) x Diacol V351 (ETO)
Patia-1 OP White La Posta (Pop. 43)
Source: CIMMYT Maize Program
tion 43 (La Posta), and Population Tuxpefio-Braquiti- mately 257,000 t of grain were produced annually on cos are important components in ICA's germplasm 185,000 ha, with national yields averaging 1.4 t/ha. complexes MB 115, MB 220, MB 233, and MB 237 CIMMYT has had a very close collaborative re(personal communication, Gonzalo Granados). search relationship with the national maize program,
Approximately 90% of the improved materials especially for the development of improved floury released in Colombia are hybrids and the balance are maize for the highland areas of the Andes (Posada open-pollinatedvarieties. Six open-pollinatedvarieties Torres 1986). CIMMYT's Andean Region Program and one varietal cross have been released in Colombia staff stationed in Cali, Colombia, have focused their that contain CIMMYT germplasm (table 14). efforts in Ecuador on the development of lowland
tropical early and full-season open-pollinated varieties ECUADOR and family hybrids (personal communication, Gonzalo
Granados). A number of improved varieties and hyMaize is the second most important cereal pro- brids have been released for commercial production duced in Ecuador, after rice. During 1983-85, approxi- (table 15).
Table 15. Maize varieties and hybrids released in Ecuador and containing CIMMYT germplasm, 1978-85
Variety name Type color Source material
Highland materials
INIAP-101 OP White Floury Pool 1, Compuesto Cacahuacintle
INIAP-130 OP Yellow Floury Highland Pop. 4
INIAP-131 OP Yellow Floury Pool 4
INIAP-180 OP Yellow Amarillo dentado duro
Lowland materials
INIAP-526 OP Yellow San Andres 7528 (Pop. 28)
INIAP-527 OP Yellow Santa Rosa 7624 (Pop. 24)
INIAP-H550 FC Yellow Fam. San Andres 7528 x INIAP-515
Source: CIMMYT Maize Program

Table 16. Open-pollinated varieties released in Peru and containing CIMMYT germplasm, 1980-85
Name color Source material
Lowland materials
PMV-748 Yellow Sete Lagoas 7728 (Pop. 28)
PMV-749 Yellow Across 7728 (Pop. 28)
Jungle areas
PMC 748 Yellow Amarillo Dentado (Pop. 28)
PMC 749 Yellow Amarillo Dentado (Pop. 28)
PMC 750 Yellow Antigua Veracruz 181 (Pop. 24)
Marginal 28 Yellow Amarillo Dentado (Pop. 28)
Highland materials
Cajamarca 10i White Floury Pool 1 (Comp. Cacahuacintle)
Cuzco 101 White Floury Pool I (Comp. Cacahuacintle)
Santa Ana 101 White Floury Pool 1 (Comp. Cacahuacintle)
Mal Paso 101 White Floury Pool 1 (Comp. Cacahuacintle)
Source: CIMMYTMaize Program
Between 1977 and 1985, CIMMYT had a maize "Sintetfco Harland" which were distributed in the early breeder stationed at the Santa Catalina Station, near 1950s. These hybrids replaced Cuban Yellow, which Quito, who was engaged in developing improved floury had previously been the major improved genotype in and morocho types for national programs in the An- use in coastal areas. dean region. This effort was very successful, and a In 1950, the Rockefeller Foundation and the Govnumber of improved highland materials have been ernment of Peru agreed to jointly collect Peruvian released in Ecuador as well as other countries. INIAP maize races and to preserve these materials in a prop101 and INIAP 130 show the most commercial accep- erly equipped germplasm bank. In 1953, the Cooperatance among farmers. The variety INIAP 526 is exten- tive Maize Improvement Program (PCIM) was offisively grown by farmers in the lowland maize areas of cially established at the National Agricultural UniverEcuador. sity at La Molina, Lima. The Rockefeller Foundation,
PERU USAID, and the World Bank have contributed funds to
PCIM, which has now been in operation for more than Maize is the most important cereal produced in 30 years. In addition to the university maize improvePeru. During 1983-85, approximately 689,000 t of grain ment program, the Ministry of Agriculture initiated a were produced on 304,000 ha, with national yields separate maize breeding program in 1980 at INIPA, averaging 2.3 t/ha. the National Agricultural Research Institute.
Ministry of Agriculture researchers began to carry CIMMYT materials are being used extensively in out intervarietal evaluation trials in the 1940s at La both the PCIM and INIPA breeding programs for the Molina Experiment Station (Paz Silva 1986). This work coastal and jungle areas (Paz Silva 1986). In particular, led to the release in 1945 of several intervarietal hy- CIMMYT Population 24 (Antigua x Veracruz 181), brids: HLM-1, HLM-2, HLM-3, NS-50, and NS-54. Population 26 (Mezcla Amarillo), Population 28 These genotypes, known as "La Molina" hybrids, were (Amarillo Dentado), and Population 36 (Cogollero) grown in the coastal areas during the late 1940s and have been especially useful (personal communication, early 1950s. Maize researchers at the Canete Experi- Gonzalo Granados). Various open-pollinated variement Station produced "Top Cross de Canete" and ties have been released by PCIM and INIPA that

Table 17. Open-pollinated varieties released in sector organizations. These hybrids involve inbreds
Venezuela and containing CIMMYT developed from the Cateto Flints, Cuban Flints, and
germplasm, 1980-85 Coastal Tropical Flints on one side and U.S. dent
inbred lines on the other (Wellhausen 1978). These Variety Grain hybrid combinations are the dark orange grain types
name color Source materials preferred by the Argentinian grain board. CIMMYT's
FUNIAP-1 Yellow Across 7328 (Pop. 28) subtropical, late yellow flint and dent populations are
used in Argentina as germplasm sources for the subFUNIAP-2 White Across 7740 (Pop. 40, QPM) tropical environments. FUNIAP-3 White Across 7523 (Pop. 23) FUNIAP-4 White Across 7843 (Pop. 43)
White Across 7822 (Pop. 22) Maize is the major cereal grown in Brazil. During
1983-85, approximately 20.6 million t of grain were Source: CIMMYT Maize Program produced annually on 11.6 million ha, with national
yields averaging 1.8 t/ha, making Brazil the second largest maize producer in the developing world, after China.
utilize CIMMYT germplasm (table 16). The newly D
released variety Marginal 28 tropical is adapted to the During the 1950s and 1960s, Brazilian maize scienjungle areas. twists made extensive use of exotic germplasm from the
Caribbean Flint and Dent complexes to form new VENEZUELA
Maize is the most important cereal produced in Venezuela. During 1983-85, approximately 645,000 t of grain were produced annually on 368,000 ha, with national yields averaging 1.8 t/ha.
CIMMYT regularly supplies germplasm to the national maize breeding program. Experimental varieties from Populations 21, 22, 29, 32, and 43 have been used to develop a broad-based white grain germplasm complex (personal communication, Gonzalo Granados). Populations 21 and 32 also have been used as a source of inbred lines. Five open-pollinated varieties containing CIMMYT germplasm have been released for commercial use by the Fundaci6n para la Investigaci6n Agricola y Promoci6n Tecnol6gica (FUNIP) (table 17).
Southern Cone, South America
Maize is the second most important cereal in Argentina, after wheat. During 1983-85, approximately 10.4 million t of grain were produced annually on 3.1 million ha, with national yields averaging 3.3 t/ha; roughly 60% of national production is exported. Approximately 2.5 million ha are located in temperate environments and the remaining 0.6 million ha in subtropical environments.
Virtually all of the maize area in Argentina is Figure 15. Trials of floury maize in Ecuador planted with single-cross hybrids produced by private highlands. Source: CIMMYT.

Table 18. Maize varieties and hybrids released by Brazil and containing CIMMYT germplasm, 1975-86
Name Type Color Germplasm source
BR-5101 OP White Tuxpefio-1 (Pop. 21)
BR-5102 OP Yellow Amarillo Dentado (Pop. 28)
Centralmex Nordeste OP White Tuxpefio-1 (Pop. 21)
BR-105 OP Yellow Suwan 1
BR-108 OP White Tuxpefio (C7) (Pop. 21)
BR-126 OP White Tuxpeifio-1 (Pop. 21)
BR-300 VC Yellow National cutivar x Suwan 1
BR-301 VC White Two CIMMYT Varieties (Closed Pedigree)
BR-302 TC Yellow Two CIMMYT Varieties (Closed Pedigree)
EMPASC 151 OP Yellow Amarillo Bajio (Pop 45.) x various temperate
EMPASC 152 OP Yellow Suwan 1
Piranao OP White Tuxpefio Braquiticos
Source: CIMMYT Maize Program.
germplasm complexes that have substantially increased programs (personal communication, Ricardo Magnathe yield potential of hybrids in Brazil (Wellhausen vaca). CNPMS may release several double-cross hy1978). Seed of the related dent populations-Azteca, brids based on CIMMYT material. In 1987, EMMaya, Piramex; Centralmex, IAC1-have all been widely BRAPA announced the release of a new hybrid, BR-201, distributed in the large region from northern Rio which is tolerant of aluminum toxicity. Two populaGrande do Sul to the Amazon basin. The Tuxpefio-based tions of quality protein maize (QPM) have been segermplasm complexes have been especially productive lected from testing 23 populations of OPM material. in crosses with the Cateto Flints and Paulista Dents Population 63-blanco Dentado-1 QPM and Popula(derived from the interhybridization of Cateto and tion 64-Blanco Dentado-2 QPM may be released for a U.S. dents). program of mixing white kernel flour with wheat flour.
The Centro Nacional de Pesquisa de Milho e Sorgo Data from ABRASEM (Associacao Brasileira dos (CNPMS) maize breedingprogram was started in 1975 Produtores de Semutes, Brazilian Association of Seed through a technical cooperation agreement with Industry) indicate that 150,000 t of commercial maize CIMMYT. The program has actively tested a broad seed was produced in 1984/85. CIMMYT derived range of pools and populations from CIMMYT, and it populations may account for 1% of total seed producis believed that in the next decade farmers in Brazil will tion. Agroceres, a Brazilian seed company, markets be using, indirectly or directly, a great many commer- 50% of all hybrid seed. International seed companies cial hybrids derived from CIMMYT germplasm. market 30% of hybrid seed, and the remaining 20% is
Fifteen populations (CIMMYT) were chosen for marketed by small seed companies. additional research and different approaches for their utilization were used. From these populations, nine CHILE open-pollinated varieties and three varietal or top-cross Maize is the third most important cereal in Chile, hybrids have been released by the national maize after wheat and barley. During 1983-85, approximately program for use in different regions of Brazil (table 668,000 t were produced annually on 129,000 ha, with 18). Selected CIMMYT populations are being used for national yields averaging 5.2 t/ha. (Average yields by inbred line development by private and government 1986-87 had reached 7.0 t/ha.)

U.S. Corn Belt hybrids are well adapted to Chilean PARAGUAY
production conditions and are grown extensively through- Maize is the major cereal produced in Paraguay, out the country. (See chapter 4). Average national and rapid growth in production has occurred during maize yields are consistently above 5 t/ha and rising, the past decade. During 1983-85, approximately 473,000 The adaption of U.S. hybrids in Chile has been good t of grain were produced annually on 390,000 ha, with enough to justify the near-complete discontinuation of national yields averaging 1.2 t/ha. maize breeding research (Venezian 1987). The Na- The national maize improvement program has
tional Agricultural Research Institute (INIA) conducts collaborated closely with CIMMYT since 1981 and has some crop management research in maize and also has evaluated many normal and quality protein maize a small research project to develop an open-pollinated pools and populations. Open-pollinated varieties currently in commercial production include Guarani V-311,
flint variety (Camelia) and for green ear corn. Some an improved version of Venezuela-i, and Guarani research on silage maize is also conducted in agricul- V-312, based on Suwan 8027 provided by CIMMYT. tural universities. CIMMYT materials are used to a The major hybrid in use is Pioneer 6875 from Brazil. A limited extent in these modest national maize breeding quality protein maize variety, Nutri-Guarani-V241, projects. has been commercially released (NRC 1988).

Given the paucity of readily available data, estimat- planted with improved OPVs and hybrids. National ing the area planted with improved maize materials program maize researchers and CIMMYT outreach and determining the origin of this germplasm are staff were sent a questionnaire asking for figures on the precarious exercises. The fact that high-yielding maize amount of seed planted and the area of maize sown materials do not have a single common characteristic- with improved genotypes. Maize seed enterprises were such as the dwarfing gene-makes the tracking of area sent a different questionnaire soliciting data on costs, data on the use of improved germplasm more difficult. s a diff er u es in dton of Moreover, since open-pollinated varieties are still the prices, yield, and other variables in the production of dominant genotypes in most developing countries and maize seed. More than 40 individuals and organizafarmer-maintained and distributed seed is the major tions responded to the questionnaires. This informaseed source, commercial records from seed sales are tion became the basis of a report recently published by not available to gauge the actual use of improved open- CIMMYT on the economics of commercial maize seed pollinated varieties (OPVs). Finally, the complex ge- production in developing countries (CIMMYT 1987). neticbackground of the new maize types developed, or Information was requested on three classes of imbeing developed, also precludes a simple data compila- proved genotypes: (1) hybrids, (2) commercial certition on the pedigrees of germplasm contained in these fled seed of OPVs, and (3) seed of improved OPVs improved genotypes. maintained by farmers for their own purposes or traded
It is likely that the data reported by national pro- among themselves. However, no attempt was made in grams on the area planted with hybrids are the most the CIMMYT study to identify the pedigrees/parentreliable, since there are more reasonable records of ages of the improved maize materials reported to be in hybrid seed sold annually to farmers and thus more use. accurate estimates can be made of the area planted The CIMMYT maize study reports that improved with these materials. The data on OPVs are much less genotypes are grown on about one-half (40 million ha) reliable, especially since in the majority of the country of the total maize area in the Third World. Thirty cases, relatively little certified seed of OPVs is sold o h ot ae areante th Worid. which through an organized seed sector. While national maize million ha of the al i World maich research and production programs have attempted to represents 38% of the total Third World maize area. introduce improved OPVs through extension demon- Hybrids are grown primarily in temperate and substrations and seed-increase plots planted on farmers' tropical production environments characterized by higher ftrtieds ad su-inrmase deliver sytems hnfavern overall input use and commercial agricultural sectors. fields, such informal seed delivery systems have not Three such countries-Argentina, Brazil, and Chinabeen able to maintain the genetic purity of the im- account for 24 million ha, or 80% of the total Third proved OPVs originally introduced, which invariably World area planted with hybrids. Improved OPVs are become contaminated by surrounding maize materials reported to be grown on some 10 million ha, which (often unimproved genotypes). represents 13% of the total maize area. This estimate
CIMMYT SURVEY OF IMPROVED includes the area planted with commerciallypurchased
MAIZE SEED USE OPVs (7%) as well as the area planted with farmermaintained seed of improved varieties (6%). Only a
During 1985-86, CIMMYT conducted a survey to few countries reported a significant area planted with estimate the total maize area in the developing world certified seed of improved OPVs, notably Thailand,

Table 19. Comparative use rates of hybrids and improved OPVs in important maize-producing developing countriesa
1983-85 1985-86
Maize area Percentage of area planted to:
000 ha Hybrids Imp. OPVs Total
Argentina 3,096 100 0 100
Brazil 11,583 63 7 70
Chile 129 68 13 81
China 18,403 72 0 72
Egypt 805 10 54 64
El Salvador 242 71 0 71
Guatemala 777 36 24 60
Kenya 1,547 61 5 66
Peru 304 7 43 50
Thailand 1,709 8 62 70
Zambia 512 53 11 64
Zimbabwe 1,450 60 17 77
Subtotal 40,557 66 7 73
Other developing
countries 38,514 11 17 28
Source: 1986 CIMMYT World Maize Facts and Trends: Thie Economics of Commercial Maize Seed
Production in Developing Countries.
aThose producing over 100,000 t of maize annually.
Egypt, and Guatemala. In these countries, effective countries reported the use of improved genotypes on OPV research programs have been linked with viable more than 50% of their total maize area (table 19). seed industries resulting in substantial sales of im- While small in number, these countries do account for proved open-pollinated varieties, about half of the total area planted with maize in the
As would be expected, greater adoption of im- developing world. In the remaining 50% of the total proved genotypes has occurred in those environments Third World maize area, spanning some 40-plus counthat are more favorable for maize production and tries, adoption of improved maize genotypes is less among those farmers more oriented toward the com- than 30%. mercial production of maize. In countries in which a While the figures in table 19 should not be taken at considerable portion of the total area is planted with face value, they do give some indication of the relative hybrids, patterns of adoption have usually followed a importance of improved open-pollinated varieties progression from local varieties to improved OPVs, compared to hybrids. Similar caution should be used next to nonconventional hybrids (varietal and family with many of the remaining tables in this section. crosses), and then to conventional hybrids (single, double, and three-way crosses). The development of REGIONAL TOTALS viable seed sectors, of course, has been a necessary condition for hybrid seed use. Regional maize production statistics and the proAt the country level, of the 54 countries that annu- portion of total maize areas planted with improved ally produce more than 100,000 t of maize, only 12 genotypes in various developing regions are shown in

Table 20. Maize statistics for developing country regions
1983-85a 1985-86b
Area Yield Prod. Percentage area planted with:
Region 000 ha t/ha 000 t Hybridsc Imp. OPVsd Total
Eastern and Southern Africae 9,769 1.2 10,520 25 11 36
West and Central Africa 5,630 0.9 5,250 1 21 22
North Africa and Middle
East 2,372 2.7 6,456 32 15 47
South Asia 7,493 1.3 10,005 11 23 34
Southeast Asia and Pacific 8,158 1.6 13,405 3 34 37
East Asia 18,905 3.7 70,766 71 1 72
Mexico, Central America,
and Caribbean 10,204 1.6 16,567 26 16 42
Andean Countries, South
America 1,761 1.7 2,923 20 9 29
Southern Cone, South
America 15,295 2.1 32,854 70 6 76
Totals 79,071 1.7 136,156 38 13 51
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Countries.
aBased on 1985 FAO Production Yearbook.
bData from survey conducted by CIMMYT in 1985-86.
Concludes conventional hybrids (single, double, and three-way crosses) and nonconventional hybrids (family crosses, variety crosses).
dlncludes seed of improved OPVs retained by farmers.
eDoes not include South Africa.
table 20. The highest rates of adoption are reported in genotypes are also offered. Summary statements are the Southern Cone countries of South America and in also made on the germplasm contributions that CIMEast Asia, regions with relatively well-developed agri- MYT and IITA (and various predecessor maize recultural systems and temperate maize-producing envi- search programs) have made and are making to naronments. The lowest rates of adoption are reported in tional maize improvement programs. West and Central Africa and the Andean countries of South America, areas where maize is grown under Eastern and Southern Africa
low-fertility conditions primarily by resource-poor This region exhibits great contrasts in the use of farmers for home consumption. Intermediate levels of improved varieties and hybrids (see table 21). For the adoption are reported in Mexico and in most countries region as a whole, hybrids account for 25% of the total of South and Southeast Asia. In these regions, adoption maize area and OPVs for about 11% of the total area. levels are lower than might be expected, especially Functioning maize seed sectors have been established when maize is compared to wheat and rice. in South Africa, Zimbabwe, Kenya, and Zambia, as
The following section reports country data on the reflected by the high proportion of the total maize area use of improved OPVs and hybrids. Brief assessments planted with hybrids. Maize seed enterprises in most of the major obstacles in increased use of improved other countries in the region are undeveloped, and

Table 21. Comparative data for Eastern and Southern Africa on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Angola 600 0.4 262 NA NA NA
Burundi 134 1.1 146 NA NA NA
Ethiopia 850 1.4 1,183 NA NA NA
Kenya 1,547 1.5 2,353 61 5 66
Madagascar 140 1.0 138 0 4 4
Malawi 1,162 1.2 1,397 9 17 26
Mozambique 600 0.6 338 NA NA NA
South Africaa 4,139 1.3 5,340 95 2 97
Somalia 138 0.9 130 0 6 6
Tanzania 1,483 1.1 1,593 5 7 12
Uganda 358 1.2 365 0 35 35
Zambia 512 1.7 863 53 11 64
Zimbabwea 1,450 1.2 1,693 60 17 77
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Counties.
aSerious drought in these countries during 1983-84 severely depressed average yields and production (up to 50% below normal trend lines).
Table 22. Comparative data for West and Central Africa on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Benin 485 0.7 370 NA NA NA
Burkina Faso 128 0.7 90 NA NA NA
Cameroon 443 1.2 510 0 30 30
Ghana 418 1.0 402 0 30 30
Ivory Coast 575 0.8 478 0 10 10
Nigeria 2,022 1.0 2,066 2 38 40
Senegal 84 1.2 101 0 30 30
Togo 179 1.1 191 3 30 33
Zaire 822 0.9 709 NA NA NA
Source: 1986 CIMMYTWorldMaize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Countries.

Table 23. Comparative data for North Africa and the Middle East on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Afghanistan 471 1.7 317 NA NA NA
Egypt 805 4.4 3,554 10 54 64
Morocco 405 0.7 267 NA NA NA
Turkey 560 2.9 1,627 33 13 46
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production.
approximately two-thirds of the total maize area is straint than inadequate germplasm. As such, the rate of planted with farmer-maintained OPV seed (local and adoption of improved genotypes will lag behind, and be improved varieties). In many countries, increased demand a function of, increased fertilizer use. for improved genotypes will be closely linked to the IITA has made significant germplasm contribuincreased use of fertilizer, since soil infertility is a tions to national programs through its streak-resistant major production constraint, materials. CIMMYT's lowland tropical white-grain
Good germplasm sources are available in most populations and experimental varieties (many of which national programs for developing improved genotypes have been converted to streak-resistant types) have for the lowland tropics. CIMMYT's white grain popu- also been extensively used by national programs; imlations have been used to develop varieties in Kenya, proved OPVs based on CIMMYT and IITA maize Tanzania, and Zambia. Streak-resistant populations materials have been released throughout the region. and OPVs developed by IITA and CIMMYT have also IITA's hybrid program has made significant contribubeen widely used in areas in which this disease is an tions in Nigeria, where newly established hybrid seed important economic problem. IITA's tropical inbred companies are now producing increasing amounts of lines and hybrids are also being used in several coun- hybrid seed based on IITA's material. tries. More germplasm development work is needed to
develop improved genotypes for the intermediate- North Africa and the Middle East elevation zones (1,000 to 1,500 m) of eastern and
southern Africa, where some 3 million ha of maize is This region (see table 23) mainly has subtropical grown. To address these germplasm development and some temperate environments. Maize production challenges, CIMMYT and IITA launched a joint breed- is significant only in Egypt and Turkey, even though ing program in 1985, in collaboration with the Univer- nearly 5 million t of maize (primarily for use as poultry sity of Zimbabwe, for the intermediate elevation areas. and livestock feed) is imported annuallyby countries in the region. Egypt and Turkey have reasonably
West and Central Africa well-developed and integrated maize research and
The region has a very low percentage of maize area seed systems in which private organizations are playing planted with improved genotypes (see table 22), and an increasingly important role. most countries have poorly developed maize seed CIMMYT's germplasm has been used extensively enterprises. Only about 1% of the region's maize area varietal development in Egypt; IITAs inbred lines is planted with hybrids. While it is reported that 20% of n variaveloen e t o Egypt. i ine the total maize area is planted with improved OPVs, at
least half of thisarea is plantedwith farmer-maintained MYT's sub-tropical populations have been used in seed. Thus, farmers rely on their own seed to plant Turkey, although the primary genetic base used in approximately 90% of the total maize area. In general, national improvement programs is temperate germplasm soil infertility is a far more serious production con- from Europe and the United States.

Table 24. Comparative data for South Asia on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Burma 185 1.7 317 0 34 34
India 5,896 1.3 7,759 13 23 36
Nepal 520 1.5 784 0 10 10
Pakistan 801 1.3 1,024 2 26 28
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Countries.
South Asia thought to be paid too little to do a quality job of seed
Maize is an important crop in the region (see table production, and the profit margins allowed seed dis24). Except in India, virtually no hybrids are used, and tributors are often so small that they are not interested most maize farmers in the region rely on their own seed in marketing maize seed. of OPVs to plant nearly 80% of the total area. The CIMMYT's tropical and subtropical materialshave relatively low percentages of the total maize area been especially useful for increasing the disease resisplanted with improved genotypes is due primarily to tance of national breeding populations that are prim arinternal inefficiencies within national maize seed sec- ily based on temperate-zone germplasm from the United tors and the general lack of market development activi- States. CIMMYT's early-maturing populations are also ties for maize and maize byproducts (poultry and being used extensively to develop shorter season varielivestock feed, sweeteners, oil). India and Pakistan ties. In recent years, national programs in the region have reasonably well-established seed production and have also requested and been supplied with inbred distribution systems that function effectively for higher lines and hybrids from IITA; these materials are still value crops such as wheat, rice, and cotton. However, being evaluated but already are showing promise. inadequate price incentives in the maize seed sector
and insufficient priority given to this crop in government development plans have resulted in relatively Adoption levels of improved genotypes vary conlittle commercial seed production of improved geno- siderably in this region (see table 25). Thailand has a types in these two countries. Maize seed producers are well-established seed sector that markets primarily
Table 25. Comparative data for Southeast Asia and the Pacific on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Indonesia 2,615 2.0 5,225 1 24 25
Philippines 3,361 1.0 3,331 1 25 26
Thailand 1,709 2.4 4,154 8 62 70
Vietnam 387 1.3 499 0 38 38
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Countries.

Table 26. Comparative data for East Asia on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
China 18,403 3.7 67,873 72 0 72
North Korea 423 6.1 2,593 NA NA NA
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Countries.
improved OPVs with downy mildew resistance, and East Asia
most farmers use these improved genotypes. In Indo- This region is composed mostly of China and acnesia and the Philippines, farmers still rely on their own counts for approximately 40% of total Third World seed to plant between 80% and 95% of the total maize
area. To date, little area is planted with hybrids though maize production (see table 26). The major maize increased rates of adoption are anticipated, especially growing areas have temperate environments where
since considerable public and private sector research is maize attains high yield levels. China's temperate zones now under way to develop hybrids. Private seed corn- U.S., and European germplasm. In China's smaller panics are increasingly active in Thailand, Indonesia, subtropical maize growing area (in southern China, and the Philippines. Rapid rates of adoption of improved gentoypes-which increasingly will be Hunan Province), CIMMYT's Tuxpefio-based popuhybrids-are expected throughout the region in the lations (normal and quality protein maize) are being coming decade. used extensively to develop new hybrids.
Scientists at Kasertsart University in Thailand, in Mexico, Central America,
collaboration with staff from the Rockefeller Founda- and the Caribbean
tion's Inter-Asian Corn Program, developed superior
OPVs with downy mildew resistance during the 1970s. Maize is the major crop in this region (see table 27). This germplasm complex (Suwan 1) has been exten- It is grown as a subsistence and a commercial crop, sively used by national programs within the region as which partly explains the highly variable pattern in the well as in national programs elsewhere. use of improved maize materials evident within the
CIMMYT's tropical germplasm, though used by region. The most effective seed sectors are found in El national programs in the region as parental material in Salvador, Guatemala, and, to a lesser extent, in Mexhybrid development programs, has been susceptible to ico. In recent years, private sector seed companies have downy mildew. Beginning in 1980, CIMMYT began a increased their activity in Mexico and Guatemala. concerted effort to introgress downy mildew resistance Improved seed sector policies and more efficient seed into three tropical lowland populations adapted to the production and distribution systems, with better linkregion. As this is achieved, CIMMYT's germplasm will ages to maize research programs, can result in markbecome increasingly useful to national maize improve- edly higher levels of adoption of improved genotypes at ment programs in the region, the farm level in the near term.
IITA has also supplied tropical hybrids and inbred Germplasm developed and distributed by CIMMYT and its predecessor organization has been used
lines to various national program, including its white extensively in lowland tropical areas of Central Amerand yellow-grain populations with combined streak ica and the Caribbean. Virtually all improved OPVs virus and downy mildew resistance. IITA's materials and hybrids released by the national programs in this are showing promise and probably will be used in new region are based on this germplasm. IITA has also hybrid combinations being developed in several coun- supplied tropical hybrids and inbreds to several natries of the region, tional programs in the region. Far less impact has

Table 27. Comparative data for Mexico, Central America, and the Caribbean on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Costa Rica 62 1.7 105 6 14 20
Cuba 67 1.3 97 NA NA NA
El Salvador 242 2.1 497 71 0 71
Guatemala 777 1.4 1,062 36 24 60
Haiti 157 0.9 147 0 1 1
Honduras 353 1.5 520 NA NA NA
Mexico 8,234 1.7 13,765 25 17 42
Nicaragua 172 1.2 213 9 8 17
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Countries.
occurred in highland areas, where traditional varieties Germplasm from CIMMYT and its predecessor are still largely grown. organizations has been extensively used by national
programs in developing improved OPVs and hybrids
Andean Countries of South America for lowland tropical areas. Between 1977 and 1985,
CIMMYT also participated in a collaborative research
Maize is the major food crop in most countries of pormwt cao' ainlMiePormt
program with Ecuador's National Maize Program to
this region (see table 28).A large part of maize produc- develop improved OPVs of floury maize, the dominant tion is produced by resource-poor farmers for home genotype in many of the higher elevation maize enviconsumption; these farmers tend to use their own seed. ronments of the Andes. Several superior OPVs have Hybrids are being grown on increasing areas in the been produced through this program and have been lowlands of Colombia, Ecuador, Peru, and Venezuela, released to highland maize farmers in several counwhere maize is produced mainly as a commercial crop. tries.
Table 28. Comparative data for the Andean countries of South America on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Bolivia 311 1.5 460 NA NA NA
Colombia 592 1.5 870 13 2 15
Ecuador 185 1.4 257 3 29 32
Peru 304 2.3 689 7 43 50
Venezuela 368 1.8 645 30 13 43
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Production in Developing Countries.

Table 29. Comparative data for the southern cone countries of South America on maize production and use of hybrids and improved varieties
1983-85 1985-86
Area Yield Prod. Percentage maize area planted with:
Country 000 ha t/ha 000 t Hybrids Imp. OPVs Total
Argentina 3,096 3.4 10,367 100 0 100
Brazil 11,583 1.8 20,638 63 7 70
Chile 129 5.2 668 68 13 81
Paraguay 390 1.2 473 NA NA NA
Uruguay 97 1.1 107 NA NA NA
Source: 1986 CIMMYT World Maize Facts and Trends: The Economics of Commercial Maize Seed Produiction in Developing Countries.
Southern Cone Countries of the diffusion of improved genotypes. Many Third World
South America maize farmers operate on the fringe of the commercial
This region (see table 29) has a well-developed agricultural sector and consume most of their harvest maize economy with many high-yielding production on the farm. Until these farmers participate more environments, notably in Argentina and Chile. Effec- actively in the commercial agricultural sector, few tive maize research and seed production systems bsd purchased inputs will be used (including improved
on ublc-pivae scto copeatin hve eenesae genotypes). In other resource-poor areas, such as in oihen puli-pivt sctr cooprivaritin have beednt much of sub-Saharan Africa, low adoption levels of lispred notypures. i varitie aruhedmnat proved genotypes also are linked to other producimroed goy es rw usre tn n hl r tion constraints, especially soil infertility; until this The ybrds gownin Agenina nd hileare more pressing yield constraint is relaxed, the diffusion based on U.S. and European germplasm. CIMMYT's of improved genotypes will be held in check. germplasm contributions have been greatest to Brazil National maize research systems must also ensure and Paraguay, countries with subtropical and tropical that they are producing technologies (improved varieproduction environments. Several tropicalx temperate ties and management practices) that are appropriate to gene pools assembled by CIMMYT have also proved the needs and circumstances of target farmers. In some to be valuable germplasm. sources for national maize caetes-led"mrv"gnoysrlaedb impaveand Chile.n epeae rasiAgn national programs have not been sufficiently superior
tina nd the farmer's traditional varieties in yield potential and yield dependability to merit adoption. In other
ACCELERATING THE USE OF cases, the "improved" genotypes have not fit well
IMPROVED MAIZE GENOTYPES within the farming system. For example, their maturity
IN THE THIRD WORLD period is wrong, or they are not suitable for the cropping associations, or their storage and food use characThe diffusion of improved maize genotypes in teristics are not acceptable. Moreover, where maize developing countries has been adversely affected by production is carried out principally by resource-poor many factors. These constraints can be classified under farmers, hybrids are often too costly a seed source even three headings: (1) those related to the physical and if they could be made available at the farm level. In such economic environment in which maize is produced, (2) areas, the best opportunity for improving the genetic those related to the focus and efficiency of the maize base of the maize materials continues to be improved research system, and (3) those related to the effective- OPVs in which farmers play major roles in seed multiness of maize seed sector. pication and maintenance.
The economic environment in which maize is pro- Finally, the maize seed sector must be able to duced in developing countries varies greatly and affects produce and deliver quality seed of appropriate geno61

types to farmers in an efficient and dependable way. In ments (Douglas 1980). First, improved gentoypes a number of countries, inadequate profit incentives developed by national programs must be appropriate (price policies) within national maize seed sectors have to the needs and circumstances of Third World maize restricted the development of effective seed produc- farmers. These varying circumstances indicate that tion and distribution systems. This is the case in South both improved OPVs and hybrids have important roles Asia, where reasonably well-developed seed sectors to play in the maize economies of most developing are operating for certain crops (wheat, rice, cotton) but countries. Second, once superior genotypes are develdo not function as well for maize. Here, the inefficien- oped, varietal evaluation and registration procedures cies in the maize seed sector are holding back increased must encourage their release and use, rather than adoption of improved genotypes. retard this process through unnecessary bureaucratic
While the organization of maize research and seed obstacles. Finally, if effective maize seed production
sectrs s anatona mater mos sucesfulsysemsthe and distribution systems are to be developed, adequate smctors isxaentiona mter mos-hv bbit sucesfu sytm ah incentives must be provided to seed producers, procesmajo exepton eingChia-hve een uil upn a sors, and distributors. Cheap-seed policies have republic-private sector partnership. While each country peatedly shown that they retard the development of must work out its own system, all successful maize viable seed sectors and thus harm, rather than help, research and seed partnerships have common ele- farmers.

PUBLISHED REFERENCES Third World Food and Feed Uses of Maize. Mexico
DF, Mexico.
Abbanur, B. 1986. Maize Improvement in Somalia. In CIMMYT. 1986. To Feed Ourselves: A Proceedings of
To Fed urslve: AProcedigs f te Frst CIMMYT. 1986. To Feed Ourselves:,4 Proceedings of
To Feed Ourselves: A Proceedings of the First the First Eastern, Central, and Southern Africa
Eastern, Central, and Southern Africa Regional Regional Maize Workshop, Held in Lusaka, ZamMaize Seed Production in Developing Countries. bia, March 10-17, 1985. Mexico DF, Mexico.
CIMMYT, Mexico DF, Mexico.0-17,198. Mexico DF, Mexico.
CIM YTMexco F, exio.CIM MY T. 1987. C1MMYI' World Maize Facts & Trends: Billing, KJ. October 1985. Zimbabwe and the CGIAR CIMMYT. 1987. C MY World Maize Facts & Trends: Report Three-An Analysis of the Economics of
Centers-.-A Study of Their Collaboration inAgricul- Commercial Maize Seed Production in Developing
tural Research. CGIAR Study Paper Number 6. Countries. Mexico DF, Mexico.
The World Bank, Washington DC.
TheWrld Bak antn Clark, JA. 1954. Preventing Extinction of Original
Brieger, F.G., J.T.A. Gurgel, E. Paterniani, A. Blumen- Strains of Corn. Nat. Acad. Sci.-Nat. Res. Counc.
schein, and M.R. Allenoi. 1958. Races ofMaize in News Report 4(5):78-81.
Brazil and Other Eastern South American Countries. Nat. Acad. Sci.-Nat. Res. Counc. Publ. 593, Clark, J. A. 1956. Collection, Preservation and UtilizaWashington, DC. tion of Indigenous Strains of Maize. Economic
Botany 10:194-200.
Brown, W.L. 1960. Races of Maize in the West Indies. Botany 10:194-200.
Nat. Acad. Sci.-Nat. Res. Counc. Publ. 792, Cutler, H.C. 1946. Races of Maize in South America.
Washington, DC. Botany Miscellaneous Leaflets. Harvard UniverBrown, W.L., and M.M. Goodman. 1977. Races of sity, Cambridge, MA. 12:257-291.
Corn. In Corn and Corn Improvement, (Ed.) G. F. Dalrymple, D.G. 1986a. Development and Spread of Sprague. American Society of Agronomy, Madi- High-Yielding Wheat Varieties in Developing Counson. pp. 49-88. tries. Bureau for Science and Technology, Agency
Chaudhry, A.R. 1984. Maize in Pakistan. Punjab Agri- for International Development. Washington, DC.
cultural Research Coordination Board and Uni- Dalrymple, D.G. 1986b. Development and Spread of versity of Agriculture, Faisalabad, Pakistan. High-Yielding Rice Varieties in Developing CounChavarriaga M.E. 1966. Mafz ETO: Una Variedad tries. Bureau for Science and Technology, Agency
Producida en Colombia. La Revista ICA, 1(1):5-30. for International Development. Washington, DC. Chibasa, W.M. 1986. Zambia Seed Company: The Debelo, A. 1986. Maize Research and Production in
Maize Seed Situation in Zambia. In To Feed Our- Africa. In To Feed Ourselves: A Proceedings of the selves: A Proceedings of the First Eastern, Central, First Eastern, Central, and Southern Africa Reand Southern Africa Regional Maize Workshop, gional Maize Workshop, Held in Lusaka, Zambia,
Held in Lusaka, Zambia, March 10-17, 1985. March 10-17, 1985. CIMMYT, Mexico DF, MexCIMMYT, Mexico DF, Mexico. ico.
CIMMYT. 1982. Guatemala's Maize Research and Douglas, J.E. (Ed.). 1980.Successfid SeedPrograms:A
Production Program. CIMMYT Today No. 14. PlanningandManagement Guide. WestviewPress,
Mexico DF, Mexico. Boulder, Colorado.
CIMMYT. 1984. CIMMYT1984 World Maize Facts & Eberhart, SA., and G.F. Sprague. 1973. A Major
Trends: Report Two-An Analysis of Changes in Cereal Project to Improve Maize, Sorghum, and

Millet Production in Africa. Agronomy Joumrnal IndianAgricultural Research Institute (IARI), All-India 65:365-373. Coordinated Maize Improvement Project. 1980.
FAO. 1986. Production Yearbook, Vol. 34. Rome, Italy. Pedigree of Experimental Hybrids, Composites, Galinat, W. C. 1977. The Origin of Corn. In Corn and Synthetic Varieties, and Multiple Crosses DevelCorn Improvement, (Ed.) G. F. Sprague. Ameri- oped and Tested by the Coordinated Maize Imcan Society of Agronomy, Madison. pp. 1-47. provement Project During 1958-79.
Galinat, W. C. 1971. The Origin of Maize. Annual Jinahyon, Sujin, 1973. Maize Germplasm Utilization in
Review of Genetics 5:447-478. Thailand. The Ninth Inter-Asian Com ImproveGhana Grains Development Project (GGDP). 1987. ment Workshop. Malaysian Agricultural Research
Eighth Annual Report, Part II Research Results. and Development Institute Kuala Lampur, Malyasia,
Kumasi, Ghana. December 10-12, 1973.
Goodman, M. M. 1978. A Brief Survey of the Races of Jinahyon, Sujin. 1979. Breeding and Genetics for Yield
Maize and Current Racial Relationships. InMaize and Quality of Corn in Thailand: A Ten-Year Breeding and Genetics, (Ed.) D. B. Walden. John Review, 1968-78. TenthAnnualReportingSession, Wiley, New York, pp. 143-158. Thailand National Corn and Sorghumn Program,
Rose Garden, Nakhon Pathon, Thailand, March
Goodman, M. M. 1984. An Evaluation and Critique of 25-27, 1979.
Current Germplasm Programs. In The 1983 Plant Johnson, C.W., and K.M. Byergo, P. Fleuret, E. SimBreedingResearch Forum: Conservation and Utili- Johnson, C.W., and K.M. Byergo, P. Fleuret, E. Szation of Exotic Germplasm to Improve Varieties. mons, and G. Wasserman. 1980.Kitale Maize: The Pioneer Hi-Bred International, Inc., Des Moines, Limits ofSuccess. Agency for International Develpp. 197-227. opment, Project Evaluation Report No. 2, Washington, DC.
Grant, U.J., W.H. Hatheway, D.H. Timothy, C. Cas- ington, DC.
salett D., and L.M. Roberts. 1963. Races ofMaize Johnson, E.C. 1974. Maize Improvement. Proceedings in Venezuela. Nat. Acad. Sci.-Nat. Res. Counc. of the Worldwide Symposium on Maize ImprovePubl. 1136, Washington, DC. mentin the 1970s. CIMMYT. Mexico DF, Mexico.
Grobman, A., W. Salhuana, and R. Sevilla, in collabo- Kaahwa, E.R., and F. Kabeere. 1986. Maize Research
ration with P. C. Mangelsdorf. 1961. Races of and Production in Uganda. In To Feed Ourselves: Maize in Peru. Nat. Acad. Sci.-Nat. Res. Counc. A Proceedings of the First Eastern, Central, and Publ. 915, Washington, DC. SouthernAfrica RegionalMaize Workshop, Held in
Lusaka Zambia, March 10-17, 1985. CIMMYT,
Hallauer, A.R., and J.B. Miranda Fo. 1981. Quantita- Mexico DF, Mexico.
tive Genetics in Maize Breeding. Iowa State Univ.
Press. Ames, pp. 375-406. Langham, D.G. 1942. Venezuela-i: Una Selecclon de
Htewy. 195,p. Rac. Mafz Recomendable. Circular No. 2. Ministerio de
Hatheway, W.H. 1957. Races of Maize in Cuba. Nat. Agricultural y Ganaderia. Dept. de Gentica, InAcad. Sci.-Nat. Res. Counc. Publ. 453, Washing- Arclua aaei.Dp.d e~ia n
tonAcad. Sci.-Nat. Res. Counc. Publ. 453, Washing- stituto Experimental de Agricultura y Zootecnia.
ton, DC. El Valle, Mexico D.F., Mexico.
Holley, R.N., and M.M. Goodman. 1986. Yield PotenHolley, R.N., and M.M. Goodman. 1986. Yield Poten- Li Jingxiong. 1987. China. Proceedings of the Second tial of Tropical Hybrid Maize Derivatives. Crop Asian Regional Maize Workshop. Jakarta and East Science 28:213-218. Asian Regional Maize Wodeshop. Jakarta and East
dcienMeo F. JnrJava, Indonesia, April 27 to May 3, 1986. CIMHomem de Melo, F. January 1986. Brazil and the MYT, Mexico DF, Mexico.
CGIAR Centers-A Study of Their Collaboration in
Agricultural Research. CGIAR Study Paper Num- Mangelsdorf, P.C. 1974. Corn:Its Origin, Evoluhttion and ber 9. The World Bank, Washington, DC. Improvement. Harvard Univ. Press, Cambridge. p.
IITA. 1986a. Solvingthe Problem ofMaize Streak Virus: 262.
A Research Breakthrough To Increase Maize Pro- Mangelsdorf, P. C., R. S. MacNeish, andW. C. Galinat.
duction in Sub-Saharan Africa. Ibadan, Nigeria. 1967. Prehistoric Wild and Cultivated Maize. In IITA. 1986b. Stable, High-Yielding Hybrids Developed The Prehistory of the Tehuacan Valley. Vol. One.
at IITA: The Key to Increased Maize Production in Environment and Subsistence. (Ed.) D.S. Byers.
Sub-Saharan Africa. Ibadan, Nigeria. Univ. of Texas Press, Austin. pp. 178-200.
IITA. 1986c. IITA 1985 Annual Report and Research Mangelsdorf, P. C., R. S. MacNeish, and W. C. Galinat.
Highlights. Ibadan, Nigeria, pp. 69-84. 1964. Domestication of Corn. Science 143:538-545.

McClintock, B., T. A. Kato Y., and A. Blumenschein. AgriculturalResearch. CGIAR Study Paper Num1981. Chromosome Constitution ofRaces ofMaize: ber 11. The World Bank, Washington, DC.
Its Significance in the Interpetation of Relationships Pray, C.E. October 1986. Agricultural Research and between Races and Varieties in the Americas. Colegio Technology Transfer in India. Report No. 1. Ecode Postgraduados, Chapingo, Edo. de Mexico, nomicDevelopment Center, University of MinneMexico. sota, St. Paul.
Mikoshiba, H. March 1971. Maize in India. Technical Pray, C.E. October 1986. Agricultural Research and Bulletin No. 2, Tropical Agricultural Research Technology Transfer by the Private Sector in the Center, Ministry of Agriculture and Forestry. Philippines. Report No. 2. Economic DevelopNishigahara 2-2-1, Kita-ku, Tokyo, Japan. ment Center, University of Minnesota, St. Paul.
NRC (National Research Council). 1988. Quality-Pro- Pray, C.E. March 1987. Private Sector Agricultural
tein Maize. National Academy Press, Washington, Research and Technology Transfer in Thailand: A DC. Preliminary Survey. New Jersey Agricultural ExNeves V., J., F. Pacheco M., N. Sanchez D., and R.D. periment Station P-02001-2-87. Cook College,
Osler. 1957. El Cultivo del Malz en el Valle de Rutgers University, New Brunswick, New Jersey.
Yaqui. Boletin 304. Oficina de Estudios Especia- Primera Reunion Centroamericana. 1954. Mejoramles, Secretariat de Agricultura y Ganaderia, Mex- iento del Maiz. Primera Memoria. Projecto Coopico. erativo Centroamericano. Turialba, Costa Rica.
Ntlhabo, P.R., and M.T. Matli. 1986. Maize Research October 24-30, 1954. (mimeo).
in Lesotho. In To Feed Ourselves:A Proceedings of Ramirez E.R., D.H. Timothy, E. Diaz B., and U.J.
the First Eastern, Central, and Southern Africa Grant, in collaboration with G.E. Nicholson C., E.
Regional Maize Workshop, Held in Lusaka, Zam- Anderson, and W.L. Brown. 1960. Races of Maize
bia, March 10-17, 1985. CIMMYT, Mexico DF, in Bolivia. Nat. Acad. Sci.-Nat. Res. Counc. Publ.
Mexico. 747, Washington, DC.
Nunes, E., and I. Sataric. 1986. Research on the Con- Reyes C., P, L.S. Wortman and E.J. Wellhausen. 1955.
straints to Maize Production in Mozambique. In Maiz Hibrido Para Tierra Caliente. Folleto de To Feed Ourselves: A Proceedings of the First Divulgacion No. 18. Oficina de Estudios EspeciaEastern, Central, and Southern Africa Regional les, Programa Cooperativa de Agricultura de la
Maize Workshop, Held in Lusaka, Zambia, March Secretariat de Agricultura y Ganaderia.
10-17, 1985. CIMMYT, Mexico DF, Mexico. Roberts, L. M., U.J. Grant, R. Ramirez E., W.H.
Ochieng, JA.W. 1986. Maize Research in Kenya: An Hatheway, and D.L. Smith, in collaboration with
Overview. In To Feed Ourselves: A Proceedings of P.C. Mangelsdorf. 1957. Races ofMaize in Colomthe First Eastern, Central, and Southern Afica bia. Nat. Acad. Sci.-Nat. Res. Counc. Pub. 510,
Regional Maize Workshop, Held in Lusaka, Zam- Washington, DC.
bia, March 10-17, 1985. CIMMYT, Mexico DF,
bia March 10-17,1985. CIMMYT, Mexico DF, Roberts, L.M., E.J. Wellhausen, G. Palacios R., and A.
Cuevas R. Rocamex V-21 and Rocamex VS-101:
Pakistan Agricultural Research Council (PARC). 1986. Nuevas Variedades Mejoradas del Maiz de Madurez
Maize Production Manual PARC Special Publica- NuvsVieasMjodsdlMazeMdrz
Maize Production. Islamabad, Pakistanal. PARC Special Publica- Precoz para el Mesa Central. Folleto de Divulgation. Islaabad, Pakistan. cion No. 7. Oficina de Estudios Especiales. ProPARC/CIMMYT. 1988. Maize Research andDevelop- grama Cooperativa de Agricultural de la Secretarment in Pakistan. PARC Special Publication, Is- iat de Agricultura y Ganaderia.
lamabad, Pakistan. ateAg a Gne a
Patifio, V.M. 1956. El Maiz Chococito. Noticia Sobre Sanchez, PA. and G.M. Scobie. Cuba and the CGIAR
su Cultivo en America Ecuatorial. America Indi- Centers-A Study of Their Collaboration inAgriculgena 16:309-346. tural Research. CGIAR Study Paper Number 14.
Paz Silva, L.J. September 1986. Peru and the CGIAR The World Bank, Washington, DC.
Centers--AStudyofTheirCollaborationinAgricul- Secretariat de Agricultura y Ganaderia. 1955. Maiz tural Research. CGIAR Study Paper Number 12. Hibrido ParaElBajioyRegiones Simnilares. Folleto
The World Bank, Washington, DC. de Divulgacion No. 19. Oficina de Estudios EspePosada Torres, R. September 1986. Ecuador and the ciales. Secretariat de Agricultura y Ganaderia,
CGIAR Centers-A Study of Their Collaboration in Mexico, pp. 9-35.

Sharma, R.P. and J.R.Anderson. October 1985. Nepal Vocke, G. December 1987. Corn Production Lags
and the CGIAR Centers-A Study of Their Collabo- Behind Use in Developing Countries. In World ration in Agricultural Research. CGIAR Study Paper Agriculture Situation and Outlook Report. WASNumber 7. The World Bank, Washington, DC. 50. U.S. Department of Agriculture, Washington,
Sprague, G.F. (ed.) 1977. Corn and Corn Improve- DC. pp. 18-25.
ment. American Society of Agronomy, Madison, Walden, D. B. (ed). 1978. Maize Breeding and GenetWisconsin. ics. John Wiley & Sons, NY, p. 794.
Sriwatanapongse, S., C. Chutkaew, S. Jampathong, B. Wellhausen, E.J. 1947. Comparacton de Variedades del
Paisuwan, C. Aekatasanawan, and S. Jinahyon. Matz Obtenidas en el Bajio, Jalisco, y en la Mesa
1985. Long-term Improvement of a Maize Comn- Central. Follecto Tecnico No. 1. Oficina de Estuposite,Suwan 1.FifthInternationalSABRAO Con- dios Especiales, Secretariat de Agricultura y
gress, Bangkok, Thailand, November 25-29, 1985. Ganaderia, Mexico D.F.
Stakman, E.C., R. Bradfield, and P.C. Mangelsdorf. Wellhausen, EJ. 1950. El Programa de Mejoramiento
1967. Campaigns Against Hunger. The Belknap del Mafz en Mexico. In La Primeria Asamblea
Press of Harvard University Press, Cambridge, Latinoamericana de Fitogenetistas. Folleto MisMass., pp. 61-71, 221-222, 255-266. celaneo No. 3, Noviembre. Oficina de Estudios
Stewart, R. September 1985. Costa Rica and the CGIAR Especiales, Secretariat de Agricultura y GanadCenters-A Study of Their Collaboration in Agricul- eria, Mexico DF, Mexico.
tural Research. CGIAR Study Paper Number 4. Wellhausen, E. J., A. Fuentes O., and A. Hernandez The World Bank, Washington, DC. Corzo, in collaboration with P. C. Mangelsdorf.
Stewart, R. September 1985. Guatemala and the CGIAR 1957. Races of Maize in Central America. Nat.
Centers-A Study of Their Collaboration inAgricul- Acad. Sci.-Nat. Res. Counc. Publ. 511, Washingtural Research. CGIAR Study Paper Number 5, ton, DC.
The World Bank, Washington, DC. Wellhausen, EJ. and L.M. Roberts. 1948. Rocamex
Timmer, C.P. (ed). 1987. The Corn Economy of Indo- V-7, Una Variedad Sobresaliente de Maiz Para
nesia. Cornell University Press. Ithaca, NY, and Sembrarse de Riego en la Mesa Central. Folleto de London, U.K. Divulgacion No. 3. Oficina de Estudios EspeciaTimothy, D.H., and M.M. Goodman. 1979. Germplasm les, Secretariat de Agricultura y Ganaderia, MexPreservation-The Basis of Future Feast or Fam- ico D.F. pp. 1-21.
ine: Genetic Resources of Maize-An Example. In Wellhausen, EJ., L.M. Roberts, and E. Hernandez X., The Plant Seed. (Eds.) J. Rubenstein, R.L. Phil- in collaboration with P.C. Mangelsdorf. 1952.
lips, C.E. Green, and B.G. Gengenbach. Aca- Races of Maize in Mexico. Bussey Institution,
demic Press, NY, pp. 171-200. Harvard Univ., Cambridge, MA.
Timothy, D.H., W.H. Hatheway, U.J. Grant, M. Torre- Wellhausen, E.J. 1978. Recent Developments in Maize
groza C., D. Sarria V., and D. Varela A. 1963. Breeding in the Tropics. In Maize Breeding and Races of Maize in Ecuador. Nat. Acad. Sci.-Nat. Genetics. (Ed.) D.B. Walden. Wiley & Sons, New
Res. Counc. Publ. 975. Washington, DC. York, NY. pp. 59-84.
Timothy, D.H., B. Pefia V., and R. Ramirez E., in, NY 598.
collaboration with W. L. Brown, and E. Anderson. Zin, K. November 1986. Burma and the CGIAR 1961. Races of Maize in Chile. Nat. Acad. Sci.-Nat. Centers-A Study of Their Collaboration inAgriculRes. Counc. Pub. 847, Washington, DC. tural Research. CGIAR Study Paper Number 19.
Vasal, S.K., A. Ortega, and S. Pandey. 1982. CIM- The World Bank, Washington, DC.
MYT's Maize Germplasm Management, Improvement, and Utilization Program. CIMMYT, Mexico CORRESPONDENCE
Venezian, E. April 1987. aile and the CGIAR Centers-A
Study of Their Collaboration in Agricultural Re- Khan Bajadur, Director of Maize Production Prosearch. CGIAR Study Paper Number 20. The gram. Rafhan Maize Products, Ltd., Faisalabad,
World Bank, Washington, DC. Pakistan. Interviewed in Pakistan, May 10-12, 1987.

Mike Barrow, Pioneer Seed Company (PTY) Limited, dence dated June 25, 1985.
P.O. Box 19, Greytown, Natal, South Africa. Cor- Ricardo Magnavaca. EMBRAPA, Centro Nacional de respondence dated October 3, 1985. Pesquisa de Milho e Sorgo, Parani, Brazil. CorreCyril E. Broderick, Assistant Professor of Plant Sci- spondence dated July 22, 1985.
ence, College ofAgriculture and Forestry, Univer- Oguti W. Mangheni, Scientific Officer-Maize Breedsity of Liberia, Monrovia, Liberia. Correspon- ing, Department of Agriculture, Kawanda Redence dated October 8, 1985. search Station, P.O. Box 7065, Kampala, Uganda.
Papa Assau Camara, Director ISRA, Secteur Centre Correspondence dated December 5, 1985.
Sud-Koolack, B.P.199, Kaolack, Senegal. Corre- Alfred Manwiller, USAID/Nairobi, Box 202, APO spondence dated June 27, 1985. New York, 09675. Correspondence dated FebruJose C. Cedefio, Facultad de Agricultura, Universidad ary 5, 1985.
de Panama, Panama, Reptibica de Panama. Cor- J.L. Marchaud, Institut de Recherches Agronomiques respondence dated June 7, 1985. et des Viveres, CIRAD-Reunion, 97487 St. Dinis
Muhammed Quasim Chatha, Coordinator, Crop Cedex,IlleDeLaReunion.Correspondencedated
Maximization Programme, Pakistan Agricultural July 9, 1985.
Research Council, Islamabad, Pakistan. (Formerly Alfred J. Moshi, Coordinator, National Maize ReNational Maize Coordinator, 1976-86). Interviewed search Programme, Illonga Agricultural Research in Pakistan, May 4, 1987. Institute, Private Mail Bag, Kilos, Tanzania. CorHarouna Dosso, Maize Program, IDESSA-CV, B.P. respondence dated October 4, 1985.
635, Bauake, C6te d'Ivoire. Correspondence dated W. Gresban Nlilane, Maize Breeder, Ministry of AgriJuly 8, 1985. culture, Chitedze Agricultural Research Station,
Yoel Efron, Director of Maize Program, IITA, Oyo Illongwe, Malawi. Correspondence dated SeptemRoad PMB 5320, Ibadan, Nigeria. Interviewed in ber 20, 1985.
Nigeria in June 1987. Mulamba Ngandu Nyindu, Director, Programme
T. Gbian. Ombella Mpako Rural Development Proj- NationalMais. Department de L'Agriculture, B.P.
ect, B.P. 1007, Banquie, Central African Republic. 3673, Lubumbashi, Zaire. Correspondence dated
Correspondence dated September 21, 1985. July 16, 1985.
G. Granados. CIMMYT Regional Maize Specialist-Asia. R.C. Olver, Head, Crop Breeding Institute, Harare
Bangkok, Thailand. (Formerly Regional Maize Research Station, P.O. Box 8100, Causeway,
Specialist-Andean countries of South America, Zimbabwe. Correspondence dated June 7, 1985.
1976-86.) Interviewed in Thailand, February 16-20, John Pali-Shikhulu, Malkerns Research Station, P.O.
1987. Box 4, Malkerns, Swaziland. Correspondence dated
Wayne L. Haag. 1984 Annual Report on Middle East September 26, 1985.
Regional Maize Program. CIMMYT. El Batan, E.R. Rhodes, Adaptive Crop Research and Extension, Mexico. Unpublished. Ministry of Agriculture and Natural Resources,
Soon K. Kim, Maize Breeder, IITA Maize Program, Private Mail Bag 540, N. Gala/Freetown, Sierra
Oyo Road, PMB 5320, Ibadan, Nigeria. Corre- Leone. Correspondence dated November 26, 1985.
spondence dated September 25, 1987. Wilfried Schwiebout, Conseiller Techniques, B.P. 5,
Enkin Kinace. National Maize Coordinator, Ministry Sotoubaria, Togo. Correspondence dated Decemof Agriculture, Ankara, Turkey. Interviewed in ber 8, 1985.
Turkey, November 19-24, 1987. R.N. Wedderburn. CIMMYT Regional Maize
Manuel M. Lantin, Chairman, Department of Agron- Specialist-Asia. Bangkok, Thailand. Interviewed
omy and Program Leader of Corn and Sorghum in Thailand, February 16-20, 1987.
Program. IPB, University of the Philippines, Los
Bafios, Laguna 3720. Personal communication, MANUSCRIPT REVIEWS
Hector J. Lizarroga. Facultad de Ciencias Agropec- MagniBjarnason, CIMMYT Maize Breeder, El Batan,
uarios, Universidad Nacional Autonoma de Nica- Mexico. (Formerly Regional Maize Specialist-West ragua. A.P. 453, Managua, Nicaragua. Correspon- Africa, 1980-85, stationed at IITA, Ibadan, Nige67

ria.) Reviewed section on West Africa during and Southeast Asia. Interviewed in Pakistan,
March 1986. November 8, 1987.
William L. Brown, Pioneer Hi-Bred International, Soon K. Kim, Maize Breeder, IITA Maize Program,
P.O. Box 316, Johnston, Iowa 50131. Reviewed Oyo Road, PMB 5320, Ibadan, Nigeria. Reviewed
first revised draft of manuscript; comments sent entire report; comments sent on September 25,
on March 24, 1986. 1987.
R.P. Cantrell, CIMMYT Maize Program Director, El Alejandro Ortega, CIMMYT Maize Breeder and LiaBatan, Mexico. Reviewed first draft of report, sion Scientist to Mexican National Maize ProMarch 1986. gram, INIFAP, El Batan, Mexico. Reviewed secD.N. Duvick, Senior Vice President-Research, Pio- tion on Mexico, March 1987.
neer Hi-Bred International, 400 Locust Street, R.D. Osler, Deputy Director General, CIMMYT, El Des Moines, Iowa 50309. Reviewed entire report; Batan, Mexico. Reviewed entire report several
comments sent on September 25, 1987. times in 1986 and 1987.
Yoel Efron, Director of Maize Program, IITA, Oyo R.L. Paliwal, CIMMYT Maize Program Associate
Road, PMB 5320, Ibadan, Nigeria. Reviewed sev- Director, El Batan, Mexico. Reviewed entire reeral drafts of the entire report; comments sent port, June 1987.
April 9, 1986, May 15, 1986, March 23, 1987, and Bobby L. Renfro, CIMMYT Maize Pathologist, El November 7,1987. Batan, Mexico. (Formerly Regional Maize
Takumi Izuno, Winrock International, MART Project, Specialist-Asia, 1980-85, and Coordinator,
Pakistan Agricultural Research Council, Islamabad, Inter-Asian Corn Program, Rockefeller FoundaPakistan. (Formerly CIMMYT Maize Breeder in tion, Bangkok, Thailand, 1970-80.) Reviewed secPakistan and Asia, 1968-79, and Maize Breeder, tions on South and Southeast Asia.
Tropical Program, PioneerHi-Bred International, S.K Vasal. CIMMYT Maize Breeder, El Batan, Mexico.
Johnston City, Iowa.) Reviewed sections on South Reviewed first draft of report, March 1986.

Appendix A
Three breeding terms are used with some frequency in this report: hybrids, synthetics, and composites. Readers not familiar with these terms may find the following general definitions of assistance.
Hybrid-A single, double, or triple cross of selected inbred lines, normally with wide variability in genetic background, that attempts to enhance certain predetermined characteristics such as yield, insect or disease resistance, stalk strength, etc., and attain hybrid vigor or heterosis.
Synthetic-An open-pollinated variety derived from the combination of a number of selected self-pollinated lines, the good combining ability of which has usually been predetermined by testing several to all possible first generation (F1) combinations.
Composite-An open-pollinated variety selected from the random combination of a large number of recognized breeding lines or accessions that in theory have good combining quality and the genetic characteristics desired for a specific location or purpose.
Synthetics and composites are generally developed for adverse or marginal maize growing conditions or where the technology, infrastructure, and demand for maize seed is not sufficient to make hybrid seed production viable.
(Adapted, with modifications, from Johnson et al. 1980.)

Appendix B
Holley, R.N., and M.M. Goodman. 1988. Yield Potential of Tropical Hybrid Maize Derivatives. Crop Science 28:213-218.
In an effort to enhance the germplasm base of the U.S. maize (Zea mays L.) crop an adaptation program was initiated in 1975. A group of nine 100% tropical hybrids were crossed in all possible combinations, without reciprocals. Progeny from each cross were sib mated for six generations with selection for earliness, low plant and ear height, tassel and silk synchronization, lodging resistance, and prolificacy. Following sib mating, selection continued through two generations of self'rg. All generations were grown in Raleigh, NC. In 1983 34 selected inbred lines were crossed onto two Corn Belt single cross testers, (A632Ht x B73) and (Mo17 x H95 x 993). The number of entries representing each cross varied greatly depending upon response to selection. The testcrosses were divided into two groups by tester and combined with eight commercial check hybrids to form two tests. The tests were grown at three locations in 1984 and 1985, with three replications of each location. Testcrosses developed from 100% tropical inbred lines crossed with elite U.S. materials were agronomically competitive with commercial U.S. hybrids. The inbreds are adapted to the southern United States and flower about 1 week later than B73. Plant height and grain moisture of the testcrosses were all within the range of the commercial checks. About 25% of the testcrosses had yields comparable to the commercial checks. The inbreds are relatively insensitive to photoperiod as a result of the apparent presence of complementary genetic systems for photoperiod sensitivity among the different tropical materials.

Bureau for Science and Technology
Agency for International Development
Washington, DC 20523