Loading...
HomeMy WebLinkAbout[005] CContributions of Conventional Plant Breeding to Food Production Norman E. Borlaug In 1979 world food production of all types reached 3.75 billion metric tons, representing 1.9 billion tons of edible dry matter. Of this dry matter tonnage, 99 percent was produced on land; only slightly more than 1 percent came from oceans and inland waters. Plant products constituted 93 percent of the human diet. The remaining 7 percent of the world's diet, animal products, also came (indi- rectly) from plants. Origin of Food Crop Species and Early Genetic Improvements We will never know with certainty when nature first began inducing genetic diversity, making recombinations, and exerting selection pressure on the pro- genitors of the plant species that would be chosen, much later, by man as his food crop species. But as the Mesolithic Age gave way to the Neolithic there Summary. Within a relatively short geological time frame, Neolithic man, or more probably woman, domesticated all the major cereal grains, legumes, and root crops that the world's people depend on for most of their calories and protein. Until very recently, crop improvement was in the hands of farmers. The cornerstones of modern plant breeding were laid by Darwin and Mendel in the late 19th century. As the knowledge of genetics, plant pathology, and entomology have grown during the 20th century, plant breeders have made enormous contributions to increased food production throughout the world. There have been major plant breeding break- throughs for maize and wheat, and promising research activities to raise yields in marginal production environments are ongoing. Since it is doubtful that significant production benefits will soon be forthcoming from the use of genetic engineering techniques with higher plants, especially polyploid species, most research funds for crop improvement should continue to be allocated for conventional plant breeding research. Archeological evidence indicates that more than 3000 species of plants have been used by man for food. Currently, the world's people largely depend on about 29 crop species for most of their calories and protein. These include eight species of cereals, which collectively supply 52 percent of the total world food calories, three "root" crops, two sugar crops, seven grain legumes, seven oil seeds, and two so-called tree food crops (bananas and coconuts). These 29 basic food crops are supplemented by about 15 major species of vegetables and a like number of fruit crop species, which sup- ply much of the vitamins and some of the minerals necessary to the human diet. Norman E. Borlaug is a former director of the Wheat Program, International Maize and Wheat Improvement Center (CIMMYT), Londres 40, Mbx- ico 6, D.F., MExico 006600. He now serves CIM- MYT as a consultant. suddenly appeared, in widely dispersed regions, the most highly successful group of plant and animal breeders that the world has ever seen-the Neolithic domesticators. Within a relatively short geological period, apparently only 20 to 30 centuries, Neolithic man, or more probably woman, domesticated all of the major cereals, grain legumes, root crops, and animal species that remain to this day as man's principal sources of food. Agriculture and animal husbandry spread rapidly from their cradles of ori- gin across vast areas of Asia, Africa, Europe, and the Americas. These migra- tory diffusions were in large part possible because of the tremendous genetic diver- sity that existed in the original land races and populations of the domesticated crop plants. This genetic variability per- mitted-with the aid of continued muta- tions, natural hybridizations, and recom- binations of genes-the spinning off of new genotypes that were suitable for growing in many environments. Golden Age of Plant Breeding Until the 19th century, crop improve- ment was in the hands of farmers who selected the seed from preferred plant types of land races or populations for subsequent sowing. By the early decades of the 1800's, a number of progressive farmers in North America were busy developing and selling superior varieties based on individual plant selections. The groundwork for genetic improve- ment of crop plant species by scientific man was laid by Darwin in his writings on the variation of life species (published in 1859) and through Mendel's discovery of the laws of inheritance (reported in 1865). While Darwin's book immediately generated a great deal of interest, discus- sion, and controversy, Mendel's discov- ery was largely ignored at first. Nearly 40 years transpired before these two strands of scientific thought were joined by Karl Correns, Erich Tschermak, and Correns De Vries in independent stud- ies. This rediscovery of Mendel's laws in 1900 provoked a tremendous scientific interest in genetics. The fact that Mendel had worked out his principles on a plant (the sweet pea) encouraged many to pre- pare themselves for a career in applied plant genetics. Methods Used in Modern Plant Breeding The three major categories of plant breeding research are divided on the basis of how various species propagate themselves. Species that reproduce sex- ually and are normally propagated by seeds-including all cereal crops, le- gumes, and most trees and shrubs-oc- cupy the first two categories. One of these includes species that set seed through self-pollination; the second, spe- cies that set seed largely through cross- pollination. The third category includes species that are asexually propagated through the planting of vegetative parts or grafting. In this article I mainly dis- cuss plant breeding achievements in wheat, a self-pollinating species, and maize, a cross-pollinating species. The cornerstones of all plant breeding are (i) the conscious introduction of ge- netic diversity into populations by inter- crossing or mating selected germ plasm with outstanding characters that comple- ment one another and (ii) the selection of superior plants with genes for desired t ~t 11 FEBRUARY 1983 689 traits until higher levels of improved adaptation (reproductive fitness), genet- ic uniformity, and agronomic stability are reached. The appropriateness of a breeding methodology is determined mainly by the sexual nature of the crop- inbreeding (self-pollinating) or outbreed- ing (cross-pollinating)-its genetic struc- ture, and the objectives to be achieved. Wheat Breeding As the knowledge of genetics and plant pathology grew during the first and second decades. of this century, wheat breeding methods evolved from bulk and pure-line selections of plants from land races to hybridization programs. With this methodology, controlled pollina- tions are made between two or more superior parent types. In subsequent segregating generations derived from these controlled crosses, the individual plants possessing the best combinations of desirable characteristics are selected and advanced to the next generation. This process is repeated until all proge- nies from an individual plant row are genotypically and phenotypically uni- form. When acceptable uniformity has been attained, the best progenies (lines) are sown in replicated multirow plots and compared with the best commercial varieties for grain yield, agronomic type, disease and insect reaction, and milling and baking quality. The trials are repeat- ed for several years at a number of different locations to obtain reliable in- formation on the interaction of the varie- ty (genotype) with different environ- ments. When a new line or variety has significantly outperformed existing com- mercial varieties over several years, it is eligible for multiplication and release as a new commercial variety. As a result of such breeding tech- niques, a number of major break- throughs in wheat breeding have oc- curred during the past four decades. Most significant among these include the progress made in plant pathology re- search to develop disease-resistant vari- eties, achievements in raising maximum genetic yield potential, and the benefits derived from broader adaptation in wheat crop cultivars. Disease resistance. The pioneering work on stem rust by E. C. Stakman in Minnesota during 1913 to 1930 revealed that the rust organism comprises a large number of pathogenic races that differ in their ability to attack wheat varieties. This discovery led to the understanding that, for a wheat variety to maintain its resistance to stem rust, it had to possess resistance to all the races present throughout the region. With this greater understanding of pathogenic organisms, breeders began to develop more stable sources of genetic resistance in different wheat cultivars. Today, many improved wheat varieties have been developed that possess far broader spectrums of polygenic resistance to many of the 30 wheat diseases that can and do cause serious economic losses in different parts of the world. Yield potential. Until about 1961 there was no significant increase in grain yield directly attributable to the increase in the maximum genetic yield potential of new varieties. The release of the first semi- dwarf winter wheat variety, Gaines, in Washington State by O. A. Vogel and his colleagues in 1961, followed by the re- lease in Mexico of two semidwarf spring wheat varieties, Pitic 62 and Penjamo 62 in 1962 and Sonora 64, Lerma Rojo 64, Super X, and Siete Cerros in 1964, changed the potential wheat yield situa- tion dramatically. The semidwarf varie- ties, all with one or two dwarfing genes derived from the Japanese winter wheat variety Norin 10, possessed a 100 per- cent yield advantage over the best previ- ously available tall commercial varieties. Compared to the taller types, semidwarf varieties have a higher tillering capacity, more grain-filled heads, and shorter stems that make them resistant to lodg- ing under higher levels of fertilization and irrigation. Perhaps even more impor- tant, however, was the change in the "harvest index" of semidwarf varieties to partitioning more of the total dry matter production to grain. As such, semidwarf varieties convert a higher per- centage of the uptake of fertilizer and soil moisture to grain than do the taller types. Adaptation. Until the 1950's, the dog- ma in plant breeding was that the only way to ensure the development of high- yielding, well-adapted varieties is to se- lect them through all the segregating generations in the location where they are to be grown commercially. Faced with the urgent need to develop accept- able stem rust-resistant varieties in Mex- ico, a decision was made to ignore dog- ma and use several ecological areas that would permit the growing and selecting of two segregating generations of proge- ny each year. With two breeding cycles every 12 months, a new variety could be developed in 4 years rather than in the 8 years required with the conventional methods. To accomplish this task in a 12- month period, we at the International Maize and Wheat Improvement Center (CIMMYT) were forced to select two very diverse environments separated from one another by 10` of latitude (with changing day lengths) and differing in elevation by 2600 meters. Segregating populations were shuttled, grown, and selected in these two very different envi- ronments. Only varieties that withstood the rigors of both environments were advanced in the breeding program. The results were startling. Not only did these varieties yield well in Mexico, they also performed well in many other environments, from Canada to Argenti- na, because of their more general insen- sitivity to differing day lengths. In con- trast, U.S. and Canadian varieties per- formed well only in the areas where they were developed. The development of these broadly adapted spring wheat vari- eties not only benefited Mexico, but later had a tremendous impact on wheat pro- duction in other parts of the world. Maize Breeding In the 19th century American farmers made important varietal improvements in the maize species by continuously reselecting the best ears from the best plants in open-pollinated varieties and regrowing them the following year. The introduction of seed from diverse maize- growing areas resulted in natural hybrid- ization with local cultivars. Natural crossing and subsequent mass selection activities gave rise to the open-pollinated dent varieties of the U.S. Corn Belt. These varieties continued in use until the development of hybrids. Development of F, hybrids. It was recognized early on that inbreeding in maize leads to reduced vigor in the fol- lowing generation and that vigor can be restored by crossing. Darwin noted this phenomenon in The Vegetable King- dom, published in 1876. The first orga- nized attempt to exploit hybrid vigor in maize was initiated by W. J. Beal at Michigan State College in 1875. Beal's work and that of others stimulated little interest for 25 years until Edward East and George Schull proved conclusively that although maize lost vigor on in- breeding, when inbred lines were crossed, the progeny of the next genera- tion exhibited an explosive recovery of vigor called heterosis. The problem that remained was how to exploit the hetero- sis commercially, since the cost of F, hybrid seed would be prohibitively ex- pensive. The solution was forthcoming from the work of Donald Jones, who had joined East's staff in 1915. In 3 years he found a solution to the high "seed cost" of pro- ducing hybrids, and in the process in- 690 SCIENCE, VOL. 219 creased yields above those of the original single-cross hybrids. His approach was to mate two single-cross hybrids, formed by intercrossing four inbred lines, to produce what is called a double-cross hybrid. This was a giant step toward solving the problem of reducing the cost of hybrid seed to the farmer. But there was no stampede to exploit this potential until the mid-1920's, when H. A. Wallace, later to become Secre- tary of Agriculture and Vice President under Roosevelt, founded Pioneer, Inc., the first private hybrid seed company. Because of the disastrous economic de- pression of the 1930's, the use of hybrids did not really take off until the early 1940's. But by the mid-1950's hybrids dominated U.S. maize production and the use of open-pollinated varieties had virtually disappeared. Since the commercial introduction of the first hybrids in the United States some 40 years ago, many improved elite hybrids with continually higher yields, improved disease and insect resistance, and shorter and stronger stalks suitable for mechanical harvesting have been de- veloped. Open-pollinated varieties. Efforts to improve maize yields in the developing world have not achieved the spectacular successes characterized by hybrid maize production on the well-watered soils of mid-America. For reasons related to the economic circumstances of farmers and the nascent agricultural infrastructure of most developing countries, the CIM- MYT breeding program has focused on developing superior open-pollinated populations and varieties to serve the special needs of Third World countries. More than 24 populations have been developed that can serve the major envi- ronmental, maturity, and grain require- ments of the developing world. For more than a decade, these populations have been improved by recurrent selection through a multilocational international testing system for yield potential, dis- ease and insect resistance, grain type, and various agronomic characteristics. Some 70 open-pollinated varieties de- rived from these populations have been released by national breeding programs in over 20 developing countries. These open-pollinated varieties are surpassing traditional Third World varieties in yield potential by 20 to 35 percent and have better agronomic characteristics and ear- lier maturity. Certain of these improved populations also maintain their yield superiority (and dependability) over a very wide range of environments. It has now become clear from the dynamic recurrent selection Table 1. Impact of improved technology on land use, crop yield, and production in the United States. Area Yield Produc- (thou- (tons tion Crop sands (thou- of hec- per sands tares) are) hect of tons) 1938 to 1940 Maize 36,014 1.80 64,104 Wheat 23,635 0.96 22,453 17 major 128,820 252,033 crops' 1958 to 1960 Maize 29,714 3.36 99,891 Wheat 21,419 1.67 15,883 17 major 127,436 391,388 crops 1978 to 1980 Maize 29,338 6.32 185,208 Wheat 25,614 2.22 57,016 17 major 132,544 610,293 crops 'Com, wheat, rice, barley, sorghum, oats, rye, cotton, soybeans, peanuts, beans, flaxseed, pota- toes, sugar beets, hay, corn silage, tobacco. program carried out at multilocational sites that it is possible to breed high- yielding, open-pollinated varieties with a broad spectrum of disease and insect resistance that can, at the same time, have an amazingly broad adaptation across many latitutdes and elevations. Improved nutritional quality. The dis- covery in 1964 at Purdue University that the mutant opaque-2 gene increases the lysine and tryptophan content of maize by more than 50 percent created consid- erable excitement among many agricul- tural scientists and nutritionists. These are the two most limiting essential amino acids in maize for monogastric animals and man. Consequently, it was visual- ized that it would soon be possible to develop high-yielding maize materials with much-improved nutritional value. However, the difficulty of this task soon became apparent. When the opaque-2 gene was incorporated in maize materi- als it brought along a host of adverse effects, including a reduction in yield of 15 to 20 percent; a dull, soft, chalky kernel; increased susceptibility to ear diseases and to insects when in storage; and a slower drying rate of grain at physiological maturity. While many maize breeders soon abandoned their work on opaque-2 maize, CIMMYT persisted in its at- tempts to develop nutritionally superior maize types with high yield potential and suitable grain and agronomic character- istics. A breakthrough was achieved with the discovery that "normal" maize populations have minor "modifier" genes that can influence the soft texture of the opaque-2 endosperm. A closely coordinated effort to overcome these de- fects was launched by two CIMMYT scientists and a biochemist. Through the development of a recurrent selection and backcrossing breeding scheme and rapid laboratory screening methods for protein quality, they were able to retain the high levels of lysine and tryptophan associat- ed with the soft-textured opaque-2 endo- sperm, while pyramiding suitable modifi- er genes to convert the soft-textured materials to normal maize types with a hard-endosperm. This conversion was done against the background of the best broadly adapted "normal" CIMMYT populations, so parallel improvements have been carried on for other characteristics simulta- neously. The most advanced of the hard- endosperm protein maize materials have been evaluated internationally at a num- ber of locations over the past 2 years. They have been found to be roughly equal in yield, in resistance to disease and insects, and in breadth of adaptation to the best normal varieties included in the yield trials. Contributions of Plant Breeding to World Food: Wheat and Maize The contributions of plant breeding research must be seen in the context of total research efforts to improve the ef- fectiveness of agricultural production. Plant breeding, or genetic improvement, is but one element in a research triad that includes improvements through more effective crop husbandry and agronomic practices as well as more productive interactions between particular environ- ments and genotypes. During the 1940's the research compo- nents needed for high-productivity agri- culture began to be applied in the United States and yield levels started their take- off, which continues today. The most spectacular increases, however, took place during the 1950's, 1960's, and 1970's with the rapid expansion of the infrastructure for the production and dis- tribution of seed, fertilizers, herbicides, pesticides, and machinery. Between 1940 and 1980 the combined production of 17 major crops in the Unit- ed States increased 242 percent, from 252 million to 610 million metric tons (Table 1). This large increase in produc- tion was obtained with an increase in the area of cultivated land of only 3 percent. Had 1940 yield levels persisted in 1980, 177 million additional hectares of good U.S. cropland would have been needed to equal the 1980 harvest. 11 FEBRUARY 1983 691 E" Table 2. Wheat production in India before and after the wheat revolution. [Data from the Indian national wheat program. Format adapted from that of B. A. Krantz] Number of adults Wheat Gross s provided with production value ' carbohydrate years (millions increase (trillions of needs by of tons) dollars) increaset (millions) 1966 to 1967 11.39 88 3 50 1968 to 1%9 18.65 1540 1970 to 1971 23.83 2576 94 1972 to 1973 24.74 2758 101 1974 to 1975 24.10 2630 % 1976 to 1977 29.08 3626 133 1978 to 1979 35.51 4912 180 1980 to 1981 36.50 5110 186 "The wheat value used is $200 per ton, similar to the landed value imported wheat in India in 1981. tCal- culations are based on the provision of 65 percent of the carbohydrate portion of a diet containing 2350 kilocalories per day, or 375 grams of wheat per person per day. The most impressive change in U.S. crop yields and production during the past 40 years has occurred with maize. Yields have increased 251 percent, due in large part to the introduction of high- yielding hybrids. A conservative esti- mate is that heterosis in hybrid maize contributed at least 20 percent to the 1980 harvest of 185 million metric tons. This was an increased production in 1980 of 37 million tons, worth approximately $4.5 billion in additional maize sales over what would have been achieved with the best open-pollinated varieties. As a re- sult of the introduction of the new maize technology, 6.7 million fewer hectares were needed for maize production in 1980 than in 1940. Major yield increases in wheat and many other crops have also been achieved in the United States. Beginning in the mid-1960's, improved agricultural technology began to reach the developing world as well. The estab- lishment of the 13 international agrictil- tural research centers over the past two decades has been a major factor in stimu- lating agricultural research on the major food crops and farming systems in the developing world. The most impressive achievements to date have been in wheat and rice. The plant breeding efforts of scientists at the International Rice Re- search Institute in the Philippines and CIMMYT in Mexico did much to avert the spectre of famine in Asia in the 1960's and 1970's. In India the introduction of high-yield= ing wheat and rice varieties, in combina- tion with improved agronomic practices that permitted these varieties to express their high genetic yield potential, has had a major impact on transforming food production. When high-yielding semidwarf varie- ties of Mexican wheat were introduced into India during 1966 to 1968, national production stood at roughly I1 million metric tons and average yields were less than l ton per hectare (Table 2). The high-yielding wheat varieties quickly took over, and by 1981 wheat production had increased to 36.5 million metric tons, largely as a result of a 100 percent im- provement in national wheat yields. The 1981 harvest increase of 25.5 million tons over the 1966 harvest represents suffi- cient additional grain to provide 186 mil- lion people with 65 percent of the carbo- hydrate portion of a diet containing 2350 kilocalories per day. Equally impressive wheat production gains have been achieved in Argentina, China, Pakistan, Turkey, and, more re- cently, in Bangladesh. Total wheat pro- duction in developing countries has more than doubled over the past two decades. Although it is difficult to quantify the individual impact of the various compo- nents of production because of their in- teractions, certainly the use of high- yielding varieties developed through conventional plant breeding research, in combination with the increasing use of fertilizers and irrigation, has been a deci- sive factor in the increasing yields. More recently, we are seeing the po- tential for major technological improve- ments in other vitally important food crops in the developing world. For ex- ample, data on maize production over the past two decades in the developing countries reveal that the average annual rate of yield increases in the 1970's was twice the rate achieved in the 1960's. This marked change, I believe, points to the beginning of a technological turning point in Third World maize production in the decade ahead. Given the importance of maize as a food and feed grain, and considering its relatively higher maxi- mum genetic yield potential among the cereals, significant productivity gains will play a pivotal role in future world food production efforts. 692 The Next Doubling: Feeding 8 Billion People World population growth dictates in large measure the increases needed in food production. Since the beginning of agriculture, world population has in- creased more than 256-fold (eight dou- blings), and now stands at approximately 4.5 billion. The challenge just to maintain already inadequate per'capita food con- sumption levels is awesome. It took roughly from 12,000 B.C. until about 1850 for world population to reach 1 billion, only 80 years to reach 2 billion, and only 45 years to reach 4 billion. We are now faced with the need to double the world food supply again by the first decades of the 21st century. The dramatic increases in yield that have occurred in American agriculture since 1940 through the introduction of science-based technology also indicate the long gestation period between the initiation of research programs and the application of results on a large scale. In the case of plant breeding, more than 50 years elapsed between initiation of the original genetic research and the time when the application of research results began to affect production significantly. The magnitude of the food production tasks ahead requires that we find ways to speed up this process of applying and diffusing research results. The significance of increased crop pro- duction in the more marginial agricultur- al areas is an especially important dimen- sion in feeding future generations. Some 600 million people live in the semiarid tropics and more than 1 billion live in tropical and subtropical areas character- ized by serious biological constraints. I must caution that agricultural research alone cannot produce miraculous im- provements in many of the more margin- al production areas. Some of the biologi- cal limitations are simply too overpower- ing for science to currently overcome. Still, we can put science to work on a number of the problems faced in margin- al land areas. Future Plant Breeding Research Priorities In some scientific circles today it is anticipated that major production bene- fits will soon be forthcoming from the use of genetic engineering. The new techniques in tissue culture, cell fusion, and DNA transfer are all being heralded as the scientific answers to increasing the breadth, level, and stability of dis- ease resistance; eliminating the need for SCIENCE. VOL. 219 conventional chemical fertilizers; and further raising the genetic yield potential of food crops. Although great progress has been made by employing genetic engineering techniques with bacteria or yeasts to increase the production of insulin and interferon, there is no firm evidence that similar results will be obtained with high- er plants, especially polyploid species such as wheat. It will probably be many years before these techniques can be successfully used to breed superior crop varieties. Furthermore, it is a mistake to assume that the transfer into crop spe- cies of disease- and insect-resistant genes through genetic engineering will result in substantially more durable vari- eties than we have been able to achieve to date. Pathogens and insects, when faced with extinction, mutate into new races capable of attacking the resistant variety. This biological reality will con- tinue to hound mankind in the years ahead. Although some research funds should be directed toward the development of genetic engineering techniques to im- prove breeding programs, I believe that most of the research funds for crop im- provement should continue to be used for conventional plant breeding re- search. There is much that remains to be done, and can be done, to further im- prove disease and insect resistance, en- hance tolerance to environmental ex- tremes, and increase genetic yield poten- tial by employing conventional plant breeding methods. At CIMMYT, increasing attention is being focused on the problems of mar- ginal production areas. Two major breeding approaches are being pursued. One involves conventional breeding pro- cedures in search of genetic variation in a particular crop species for added toler- ance or resistance to specific agroclimat- ic and soil stress conditions. Improved genetic materials-in terms of drought, cold, and heat resistance and tolerance to mineral toxicities such as those found in saline and acidic soils-are emerging from this work. As one example, wheat researchers have identified materials with significantly greater tolerance to acid soils characterized by aluminum toxicity. Aluminum-tolerant wheat lines, developed in cooperation with Brazilian scientists, are showing extraordinarily high yield levels under this soil-stress situation. There are millions of hectares of potential wheat land with acid soils high in soluble aluminum that now can be brought into much higher yielding production. Wide crosses between plant species are also being explored to transfer useful genes for added environmental stability in major crop species. Triticale-a hy- brid of wheat and rye-is an example of research efforts that led to the develop- ment of a new crop species. In just two decades, tremendous strides have been made in increasing yield potential and improving agronomic types of triticale. Triticale yields have doubled and now are similar to those of the best bread wheats in optimum production environ- ments. Triticale probably has a higher genetic grain yield potential than bread wheat because of its greater production of dry matter. Its strong production ad- vantage over wheat is most evident in certain marginal areas characterized by cool temperatures, acid or sandy soils, and heavy disease pressure. In such en- vironments triticale has shown a sub- stantially higher yield advantage over wheat. Research to cross domesticated spe- cies with related wild species is another promising research avenue that may lead to the development of varieties with greater yield potential and dependability in a number of important marginal areas. Generally, such wide crosses involve the breaking down of natural barriers be- tween plant species in order to introduce useful genes from alien genera into do- mesticated crop species. We have identi- fied a number of wild species with great- er resistance to certain diseases and in- sects and tolerance to salinity, tempera- ture, and moisture stresses than we have found to date in the germ plasm of the major crop species. Successful introgres- sion of these desirable genes can lead to crop varieties with greater tolerance to environmental stresses. I am convinced that the 8 billion peo- ple projected to be living 40 to 50 years from now will continue to find most of their sustenance from the same plant species that supply most of our food needs now. Fortunately, we still have large amounts of exploitable yield poten- tial on which to capitalize, especially in the developing world, where eight of every ten new births occur. It is in these areas of the world where it is imperative to close the gap between actual and potential crop yields. We must also con- tinue to work aggressively to raise aver- age yields in the developed nations. Such yield increases will be more difficult to achieve as the maximum genetic yield of each crop species is approached. New techniques, such as tissue culture and genetic engineering, offer potentially great payoffs and merit research re- sources in the years ahead. However, we should not neglect the more convention- al areas of plant breeding research, since they represent the major line of defense today on the food front. 11 FEBRUARY 1983 693