Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 Past Experience with Genetic Modification of Plants and Their Introcluction into the Environment For thousands of years, plants have been improved by genetic mollification. Ancient agriculturists selected plants with desirable traits from landraces of domesticated relatives of wild species. Lalld- race populations consist of mixtures of genetically different plants, Al of which are reasonably adapted to the region In which they evolved but differ ~ many characteristics including reaction to disease and insect pests. With the rediscovery In 1900 of Mendel's concepts of mberitance, the scientific application of genetic principles to crop im- provement began. Each scientific advance has increased our ability to alter the genetic makeup of plants predictably, and several tech- niques are often used together to improve plants. For example, an existing plant chosen for genetic modification by recombinant DNA techniques knight have been modified by many generations of cIassi- cat breeding and selection; the recombinant plant derived from the Original could then be reintroduced into a classical breeding program from which its descendants would be released for commercial use. Each technique for genetic modification constitutes only one compo- nent in the entire crop-improvement process. Figure ~1 indicates the sequence of scientific advances that has given us our present ability to modify plant genomes in ways and at a pace heretofore impossible. The basic goal of improving crops and other plants, which is still berg pursued actively, includes unprovement of agronomic traits, cro~end-use quality, and pest resistance. 16
17 CO ~ by Z . i_ ~ ° z Z ~o ~ Z O ~ - | G o LLl C) =0 o o _ I ~ JZ C) F_Z ~ ~ UJ Z `] ~ ~ $ ~ O Yl ~ ~ : UJ ~O m \ ~ S \ P4 O . ~ _ ~m a' 0 ° Ed 4- ~ Z O O C) ~ l1J ~C' Z O
18 In this chapter we place the remaining chapters in perspective. This chapter includes discussions of the classification of various tech- niques for genetic modification, the results of genetic modifications, case studies of field introductions of crop plants, and our experience with confinement methods. TYPES OF GENETIC MODIFICATION IN PLANTS The many techniques available to modify plants genetically can be divided into three main categories: classical, cellular, and molec- ular. Each of these results in genetic variation, but each provides a different avenue for producing a plant with desirable traits. Classical Techniques of Genetic Modification of Plants Hybridization. Most genetic modification techniques are used by plant breeders whose purpose is to apply the techniques to im- prove plants with commercial value. Historically, breeders have been limited by the natural or induced sexual compatibility of plants to be hybridized in their cro~improvement programs. However, new techniques, such as molecular techniques for genetic modification, are used in crow and other plant-improvement programs to bypass the sexual hybridization step. These newer techniques complement those of classical plant hybridization. Undirected Mutagenesis. Mutations can be induced in the DNA of plant cells by such techniques as the use of DNA-altering chemicals or ionizing radiation (x rays). Intact plants or plant cells are treated with the mutagenic agent and then selected for desirable traits. This process is random, and it can induce undesirable as well as desirable changes. Mutagenesis has been used effectively to generate agricul- turally important traits (Konzak et al., 1984~. Although the range of useful variations has been narrow, more than 150 plant varieties bearing traits induced by mutagenesis have been released. Anther and Ovate C?titure. In plant breeding and in other plant research, it is sometimes desirable to have plants with half the original number of chromosomes. If a plant is diploid (2x), haploid (1x) gametes or cells found in the anthers and ovules can be cultured to produce haploid plants. These genetically modified plants can then be used in breeding or in basic research. Anther and ovule culture used for obtaining haploids is followed by chromosome doubling to
19 give homozygous diploid plants for use as cultivars or as parents of hybrids (Chase, 1969~. Embryo Rescue. Embryo rescue or culture is a procedure where- by a sexual cross yielding a viable embryo but abnormal endosperm is "rescued by culturing the embryo from the nonviable seed to produce a mature plant. This cultured plant can be used ~ further breeding; for example, the procedure has been used as an integral part of producing barley varieties (Choo et al., 1985~. Cellular Techniques of Genetic Modification of Plants Somaclonal Variation. Somaclonal variation occurs in plants regenerated from cell in tissue culture, presumably as a result of stress imposed on the plant cells. The genetic changes underlying somaclonal variation include whole chromosome changes, small and large deletions and chromosome rearrangements, single base changes, and insertion mutations resulting from the activation of cryptic trans- posable elements (Orson, 1983; Vasil, 1986~. Cell Fusion. As ire sexual hybridization in breeding, cell-fusion techniques recombine plant genomes. Cell fusion is especially useful with plants not fully sexually compatible. The cells are dissociated from tissues, walls are stripped from the cells, the membranes of the resulting protoplast are modified to facilitate fusion, and after fusion the protoplasts are cultured and regenerated into intact plants. This technique can produce novel combinations of nuclei, mitochondria, and chIoroplasts (Ehlenfel~t and Helgeson, 1987~. Molecular Techniques of Genetic Modification of Plants Molecular techniques offer several advantages and complement existing breeding efforts by increasing the diversity of genes and germ plasm available for incorporation into crops and by shortening the tune period for commercial release. The many molecular techniques for genetic modification of plants can be divided into two main types: vectored and nonvectored. These techniques are discussed in detail in Chapter 5. Vectored Modifications. Vectored notifications rely on the use of biologically active agents, such as plasmas and viruses, that facil- itate the entry of the foreign gene into the plant cell.
20 Nonvectored Modifications. Nonvectored modifications rely on the foreign genes being physically inserted into the plant cell by such methods as electroporation, rn~croinjection, or particle guns. TH1: RESIJ[TS OF GEN1:TIC MODIFICATION Plant breeding has sought to make two major kinds of modifi- cations In recipient organisms: those to increase yield and those to increase reliability of performance. Increased Yield and Increased Reliability of Performance Maize breeders have looked for varieties or hybrids that produce larger amounts of grain per unit of land area, potato breeders for increased tuber yields, and cotton breeders for increased yields of lint (fiber). In addition to breeding for greater yield one may breed for a product with more desirable qualities. Breeders of bread wheats, for example, must combine selection for maximum yield with selection for an optimal balance of the endosperm proteins required for good bread-making. Cotton breeders must select for maximum yield of fiber that also has desirable spinning characteristics. The second obligation of plant breeders has been to select for reliability of performance. Components of reliability include resis- tance to diseases and pests as well as with the physical environment. Varieties that produce bumper yields In favorable growing seasons but fail to produce a crop in unfavorable seasons cannot be accepted by subsistence farmers. Their livelihood each year depends on the crops produced ~ the previous year. Commercial farmers In today's industrial nations have a less stringent requirement for reliability because storage facilities, crop insurance, and government subsidies reduce some of the problems caused by seasonal Inconsistencies in production. But in the long run, commercial farmers need reliabil- ity of performance as well. Thus, plant breeders select for reliable varieties able to produce high yields of good quality. Changes in Plant Architecture. Plant breeders, in modifying plant varieties, have selected them for their ability to produce changed and often highly unbalanced proportions of seeds, tubers, leaves, or whichever specific plant part is of economic or aesthetic interest. Genetically modifying an organism to increase the proportion of a specific plant part nearly always reduces the ability of the organism to
21 maintain itself in the wild. Maize is one of the best known examples of a highly productive cultivated plant that cannot reproduce itself without human assistance. Its large, naked seeds bound together in a large ear and having no dispersal mechanism are notoriously id-adapted for survival ~ the wild. Changes in Pest and; Disease Resistance. Plant varieties have been continually selected for Unproved resistance or tolerance to external factors that inhibit their inherent productivity. They have been selected for resistance to insect pests, to disease organisms, and, in recent years, even to specific herbicides. If such Unproved cultivars were also able to persist in the wild, they presumably would be better adapted (at least in the short term) to persist in the presence of clisease, insects, and herbicides. Improved Tolerance to Environmental Stresses. Cultivated plant varieties have also been selected through the years for bet- ter tolerance of environmental constraints to growth. Improvements are made In, for example, heat and drought tolerance, ability to withstand] high moisture, tolerance of cold, ability to withstand ex- cessive salts or high aluminum content ~ soils, ability to withstand iron deficiency Educed by excessive alkalinity, and ability to prevail in competition with weeds through quick germination and extremely rapid growth in the seedling stage. If such unproved cultivars per- sisted in the wild, they presumably would be better adapted to survive ~ the presence of a number of environmental constrmots to growth. Breeders have a long history of incorporating these types of traits into crops without any evidence of enhanced weediness. MODIFICATIONS AND THEIR l0Fl~lDCTS ON PERSISTENCE Although domesticated plants in general cannot survive and re- produce unless aided by humans, different degrees of survivability are found among different crops and at various levels of domesti- cation within a crop. Further, genes from domesticates} plants can potentially be transferred ~ pollen from these plants to their wild relatives. Thus, whether a cultivated crop is closely related to in- digenous wall relatives Is a factor that can affect survival of at least some of the genes or gene linkage blocks of domesticated plants.
22 Degree of Domestication Maize has been cited as a cultigen so highly domesticated that it cannot survive and spread on its own. At the other extreme is a crop like Cuphea, just now being domesticated for use as an oilseed crop. Breeders have not been able to alter C?`phea's self-sow~ng nature- the seeds drop from the plant at maturity, as ~ the wild species (Knapp, 1988~. Thus, cultivated Cuphea could easily revert to the self-perpetuating nature of the wild species if other plant traits have not been altered by domestication to hincler survivability. Most of the widely grown grain crops and the horticultural and vegetable crops are at the maize end of the reproductive spectrum; they cannot survive in the wild. Many of the forage and pasture crops alfalfa, cool-season and warm-season grasses cluster nearer the other end; they can persist with some degree of success or even to- tal success. Each crop needs to be considered on its own capabilities for persistence and self-reproduction. Both the level of clomestica- tion and the reproductive phenotype of the plant must be considered. Thus, a highly selected hay or pasture crop, well-suited for farming needs as a forage plant, may be virtually unselected for any change in its seed dispersal mechanisms or ~ the ability of its seed to sur- vive and give viable seedlings in the wild. Most alfalfa varieties, for example, still have a strong tendency to produce seed in dehiscent (seif-sow~ng) pods, and seed dormancy may allow it to lie In the ground for years before germinating. Selection in alfalfa has been primarily for disease resistance and altered plant habit for chang- ing the phenotype of stem and leaf not for altered reproductive structures. Plant Habit Plant architecture has a great effect on persistence and repros auction. The bush nature of the common garden bean greatly limits its adaptability; the wild bean ~ Mexico Is a climbing vine, well- suited to survival by ctirnbing up to sun and air on stems of sturdy tall grasses such as teos~nte. In contrast, selections of Asian grass (SOT9haSt,Um Titans), a highly vigorous and desirable United States warm-season pasture grass, are unchanged in plant phenotype from their wild prairie progenitor. These cultigens might be more compet- itive than their unselected progenitors if they were introduced back into native prairie ecosystems since they have been selected pr~rnariTy for vegetative vigor.
23 Grain, vegetable, and fruit crops are generally selected for highly modified plant habit or fruit type that would not be favorable to persistence in the wild state; forage and pasture crops tend to differ less from wild relatives, but even they may have a more upright plant habit and faster growth rate. Such changes might place them at competitive disadvantage over tone ~ the struggle for survival in the wild. Adaptability, Range of Habitats Survivability in the wild can be a broad-ranging but ill-defined term. The wild environment can refer (1) to pristine natural stands of vegetation essentially unaltered by humans or (2) to untended vegetation that is nevertheless altered by human activity because of such practices as lumbering, slash-and-burn agriculture, pasturing, or incidental traffic. Or the term can refer simply (3) to survival of ~wild" plants weeds In cultivated fields. In general, domesti- cated plants have closest affinities to wild plants adapted to growth in periodically disturbed habitats. One theory contends that most domesticated plants were selected from the class of plants we now call weeds plants well adapted to be pioneers, that is, rapid invaders of patches of ground laid bare by natural phenomena such as wind, fire, or flood (Anderson 1952~. Humans with hoes, spades, and fire reproduced nature's open spaces In order to aid or ensure the growth of certain desired species already adapted to such conditions. Other unwanted pioneer species were thereby encouraged unintentionally, and came to be known as weeds. Domesticated plants and their weeds have thus evolved together, and distinctions between them are sometimes minor. For example, grassy annual sorghums, grown as pasture crops or for cutting as green forage, have often retained their wild ancestors' traits of bear- ing self-sowing, long-lived seeds with varying periods of dormancy. Thus, they are adapted to selection for survival and reproduction as weeds in row-crops such as maize, where they can grow to maturity. Such revertant forage sorghums [known to farmers as shatter-cane, (Chapter 4~] have a further preadaptation to the modern chemical age. They have the same general pattern of herbicide resistance as maize (a fairly close relative taxonorn~cally) and so are not controlled by most corn-field herbicides. Shatter-cane, in areas like Nebraska where a typical rotation is maize to sorghum, has become a weed;
24 it ~ controlled through the use of herbicides, cultivation, and crop rotation (Nilson et al., 1988~. Thus, range of adaptation to soil, water, climate, and chemicals is important in determining possible persistence of a cultigen. CASE STUDIES OF INTRODUCED CROPS When exotic plant species (wild or domesticated) are introduced into a new geographic location, their adaptability is uncertain. The vast majority of introduced species fad] to establish populations that result in significant environmental harm (S~mberIoff, 1985~. Most crop introductions (domesticated exotic species, such as soybean) have provided a large societal benefit and have caused either no or only very localized problems. A few plant introductions (usually exotic species, such as ku~zu) have established themselves as weeds. The vast majority of the crop plants grown in the United States have foreign origins. Only a small number of crops including sun- flower, cranberry, Jerusalem artichoke, blueberry, and strawberry originated here. The bulk of the a~ric~,lt~,ra1 nr~rl,,~t.i~n in t.h" United States has depended on the introduction of exotic species such as wheat, soybeans, peaches, cherries, apples, tomatoes, pota- toes, and peas. This can be an inconvenience for breeders, because the useful gene pool found in wild relatives may be less readily ac- cessible. This also can be an advantage, as genes introduced into these crop plants are not likely to spread to wild weedy populations because the growing area does not harbor native cross-hybridizing species. Instances ~ which introduced crops have escaped cultivation and have become localized weed problems are rare (see Chapter 4~. _ _ ~_ _ _ ~~ ~ _ ~v ~ _ Soybean The genus Glycine can be divided into two subgenera, which appear to have different geographic origins. The subgenus Glycine is distributed predominantly in Australia, and the subgenus Sofa primarily in China and adjacent are=. The cultigen (cultivated soybeans, Glycine man (lo.) Merr., is in the subgenus Sofa and originated genetically in China. The gene pool for the cultigen is limited to its relatives in the subgenus Sofa, as only limited success has been achieved in hybridizing the cultigen with species ~ the subgenus Glycine (Hymowitz and Newell, 1981~. Between 1765 and 189S, the soybean was introduced into the
25 United States on many occasions and was grown both in small plant- ings and commercially for hay and as a forage crop. In 189S, only about eight cultivars were grown in the United States. However, a 1928 collecting trip to Japan, Korea, and northeast China brought back 4,451 new accessions to the United States (Hymowitz, 1984~. Evaluated in field plantings throughout the country, these acquisi- tions contained a high degree of genetic variability that would be useful to breeders; for example, the genes carried resistance to many damaging diseases, such as brown spot, purple seed stain, Phytoph- thora root rot, soybean mosaic, and root-knot nematode (Hymowitz, 1984~. The soybean has been genetically modified with Agrobacterium- based transformation techniques (Hinchee et al., 1988) and with particle-gun technology (McCabe et al., 1988~. These methods stably integrated the DNA in the soybean chromosomes. These methods have produced herbicide-tolerant soybeans, and field tests are being planted ~ the United States in 1989. Extensive breeding programs have allowed the United States to become a world leader, producing 56 percent of the worId's soybeans in 1985 (Hymow~tz, 19873. Soybeans are grown on about 65 ganglion acres of farm land annuaDy in this country (USDA, 1986) and are a vital part of the nation's farm economy. Canola Canola is the general term for rapeseed In the genus Brass~ca developed by Canadian plant breeders in the 1950s to 1980s (Dc~wney and Rakow, 1987~. Historically rapeseed of] has been used as a lubricant and as an edible oil. The need for marine lubricating oils during the Second World War motivated Canadian farmers to initiate commercial growing of rapeseed, but the need disappeared after the war and production declined. Experiments In the 1940s and 1950s demonstrated that erucic acid, one of the major fatty acids in rapeseed oil, Is metabolized poorly by mammals. ~ addition, erucic acid, when fed to test animals ~ sufficient quantities, was shown to induce heart lesions. Another drawback was that the meal recovered after of! extraction was limited as feed for nonruminant animals because of its high level of glucos~nolates, compounds that release goiterogenic agents after enzymatic hydrolysis. By classical plan~breeding methods, Canadian scientists se- lected variants and produced varieties with low concentrations of
26 erucic acid in rapeseed oil (called LEAR oil), and they were released for commercial production In the late 1960s. A Polish cultivar of Brassica napes was identified with low glucosinolates, and this char- acteristic was rapidly introduced into LEAR. ~Double-Iow~ rapeseed varieties (low in erucic acid and glucosinolates) were released in 1974 in Canada and are now being introduced into Europe. The acreage of rapeseed in Canada in~r~z~PA Her ~r;+t~ ~_1~ ~$ developments. Rapeseed' including canola, is sensitive to herbicides, making weed control difficult. In addition, atrazine soil residues make it diffi- cult to grow rapeseed in fields treated with atrazine. In the late 1970s and early 1980s, plant breeders incorporated atraz~e resistance Dom certain native Brassica weedy species into canola. A 20 percent re- duction ~ yield is associated with herbicide resistance; however, more recent atrazine-resistant canolas show less yield penalty. Using molecular techniques, scientists have now produced a glyphosate- tolerant canola that has been field-tested in Canada (R. K. Downey, Agriculture Canada, personal communication, 1989~. The double-low Brassica napes and B. campestris varieties were the first rapeseed to meet specific quality requirements of low erucic acid and low glucosinolates. Rapeseed oil must contain less than 2 percent erucic acid, and the solid component of the seed must con- ta~n less them 30 m~cromoles of glucosinolate per gram to be classified as canola. Canola is now being adopted as a crop internationally. Canola oil was designated GRAS (generally regarded as safe) in the United States as LEAR oil in 1985 and as canola oil in 1988. Canola of] has become the major edible oil ~ Canada, and its use world- wicle is growing. Oilseed rape can be transformer} by Agrobacterium vectors (Fry et al., 1987) and may represent one of the first crops in which herbicide and disease-resistant plants produced by molecular modification are commercialized. ~ -~-~J ~ Bus ~ ~1 alla; auk Potato The early stages of domestication of the potato occurred about 8,000 years ago in the altiplano region of the border between Peru and Bolivia. It first appeared in Europe during the latter sixteenth century (about 1570 in Spain and 1590 in England). Potatoes were introduced into Germany, Poland, and Russia by the end of the sev- enteenth century and were of great commercial importance by the
27 second half of the eighteenth century. They were brought from Eng- land and Ireland to North America between 1620 and 1680 (Hawkes, 1982~. The potato is unexcelled among cultivated plants In the abun- dance of related germplasm and the ease of incorporating this germ- plasm into cultivated forms. About 180 tuber-bearing wall species and several primitive cultivated species are known. They are dis- tributed from the southern United States to southern Chile, with the largest number of species in the Andean regions of Peru and Bolivia. Potatoes occur from sea level to an elevation of more than 4,000 me- ters and in nearly every type of ecological location. They represent a polyploid series from diploids to hexaploids (Hawkes, 1982~. Most important, the primitive cultivated and wild species are indispens- able sources of resistance to diseases, pests, frost, and drought as wed as sources of valuable processing characteristics. They also rep- resent significant genetic diversity for breeding for heterotic (highly heterozygous) genotypes. Resistance to viruses, bacteria, fungi, nematodes, and insects has been identified in primitive cultivated and wild species. Resistance has been successfully incorporated into useful cultivars by hybridiza- tion and selection. The extensive efforts to breed for disease and pest resistance, particularly In Europe, have led to the incorporation of germplasm from several species into many cultivars. The majority of cultivars in Europe and North America contain germplasm of from one to six species. Genes of Solanum demissum (a hexaploid species from Mexico with blight resistance) are incorporated into more than 50 percent of all cultivars. The genetic diversity provided by S. demissum benefits yield (Ross, 1986~. It has been possible to hybridize aIrnost all wild species to the common cultivated potato either directly or indirectly by use of multiple crosses. Through several backcrosses of hybrids to existing cultivars, new, acceptable cultivars were obtained that contain the desired germplasm from the wild species. No undesirable ~wild" trait has been observed that has not disappeared during this procedure. From the tone of early domestication of the potato to the present, thousands of cultivars have been bred and released, and several hundred of these have been grown on large acreages. Other plant species and the environment have apparently suffered adverse effects. One cultivar, found to have unsafe levels of particular alkaloids in the tubers, was withdrawn from the market. Advanced selections are now required for alkaloids to be tested before they are released (as
28 required by GRAS) to ensure their acceptability in this regard. The variety with high alkaloids Is the parent of several other important varieties, aD of which have low levels of alkaloids themselves. The potato Is a favorable organism for cellular and molecular manipulations for two reasons: (1) Plants can be regenerated from protoplasts, leaf cell clusters, caDi, and organized tissues such as stem apical meristems, and (2) Agrobacterium Ti-plasmids can be used for transformation (Fraley et al., 1986; Ooms et al., 1987~. The direct transfer of genes for resistance into highly developed cultivars with gene-transfer methods would be significantly more effective than if done by classical breeding. Potato plants regenerated from protoplasts or other unorganized groups of cells display an outburst of phenotypic variation. Some of this somaclonal variation is due to chromosomal changes, but the basis of other variation is not known. However, the somaclonal variants resemble the variants fount] in progeny from sexual crosses. Somaclones with an unproved specific trait have been identified, although their overall performance has not been superior to the parental clone (Ross, 1986~. Somatic hybrids have been generated from both intraspecific and interspecific cell fusions. Many fusion hybrids between 24- chromosome Solanum tuberosum clones and the sexually incompat- ible, wild non-tuber-bearing species Solanum brev']ens have been produced. These hybrids are of particular interest, since some are resistant to potato leaf roD virus (Austin et at., 1985; Gibson et al., 1988~. Although chromosome number varied among the hybrids, sev- eral had the expected 48 chromosomes. Further, these hybrids can be hybridized to cultivars to obtain progeny for further selection and evaluation. A wide range of phenotypic variation among the somatic fusion hybrids resembled the somaclonal variation found in plants re- generated Tom protoplasts. Through special crosses, germplasm of S. brevidens can be incorporated into S. tuberosum by sexual crosses. The products of cell fusion are phenotypically similar to those of these sexual crosses. Maize (Corn) Introduction of new maize varieties into new environments prom ably has occurred since maize was first domesticated in Mexico, several thousand years ago. Maize entered North America several hundred years ago, constantly selected by Native Americans to allow
29 its adaptation to northern climes and varying disease and weather problems. Only inferences, based on the archaeological record, sug- gest actual events in early times. However, events of the past 50 years are reasonably well detailed and documented. Since the 1930s, maize breeders have relied on sexual crossing of elite, highly developed breeding lines followed by genetic recombi- nation during several generations of self-pollination to develop new inbred Imes that are suitable parents of commercial maize hybrids. The next step, yield testing for a 3- to 5-year period in both small plots and on those as large as farms, is crucial to developing seed products and to identifying new commercial hybrids with stable per- formance across a number of growing environments. Gene flow from commercial maize varieties to the closely related teosintes in Mexico has been studied (Smith et al., 1981~. Annual testes (closely related to maize, and also considered interfertile with it) exist in Mexico as weeds in corn fields and as completely wild species. For thousands of years, farmers in Mexico have been selecting specific new varieties of maize and reproducing them under conditions that allow the maize pollen to faU freely on stigmas of teosinte plants growing in the maize fields or nearby. Thus, there has been ample opportunity for the farmers' Deliberate release" to spread maize genes into the teosinte populations. Maize is notorious for being unable to persist in the wall because its seeds are unpro- tected and are tightly bound together In large ears, thus preventing their dispersal. Contamination of teos~tes with maize genes for these traits would decrease the ability of the teosintes to persist ~ the wild. Nevertheless, various types of teosinte have maintained their dist~c- tive phenotypes and their ability to reproduce and persist in the wall (Doebley, 1984~. Biologists believe that there is lirn~ted gene flow from maize to the teosintes (and from teosintes to maize), but such gene flow does not seem to be detrimental to the teosintes nor to change their basic nature as distinctive wild races and species. For decades, corn breeders have been modifying the corn genome by conventional breeding methods. Two situations are discussed here to exemplify the type of problems that have developed and how they have been readily managed by plant breeders. The first example is breeding for resistance to northern corn leaf blight fungus (XeZminthosporium turcica). A major gene for resistance to northern corn leaf blight, called Ott, was introduced from two sources into U.S. corn-belt breeding populations about 25 years ago. It was bred into important inbred Imes and widely used in
30 hybrids. For many years the gene provided useful degrees of tolerance to northern corn leaf blight. Recent years have seen the appearance of a new biotype of the disease organism that flourishes in maize plants containing the Htl gene. Thus the protection afforded by Htl against the disease wan greatly reduced. Because U.S. maize breeders had routinely and continually bred with non-Ht] sources of resistance to northern corn leaf blight, new hybrids were available immediately to substitute for those that suffered from the new race of northern corn leaf blight (D. N. Duvick, Pioneer Hi-Bred International, Inc., personal communication, 19893. The second example is that of the southern corn leaf blight epi- dern~c. In 1970, approximately 15 percent of the U.S. corn crop was destroyed by the fungal plant pathogen Helminthosporium maydis, which causes southern corn leaf blight (Zadoks and Schein, 1979~. This represented a loss of 20 million metric tons of corn, worth about one billion dollars. Southern corn leaf blight was not a new corn disease, but, rather, one that had been controlled successfully with a variety of resistance genes. What then could account for the problem in 1970? Two key factors were involved: the natural development of a new race of the pathogen, race T. and the extensive use of hybrid Imes with Texas cytoplasm~c male sterility, Tome. The first factor to consider is the development of Belminthospo- rium maydis race T. Plant pathogens are continually evolving in re- sponse to selective pressures from changes in their environment, such as the introduction of new types of host plant resistance genes. This usually yields a number of different races that may be isolated geo- graphically or biologically on more suitable alternative host plants. This was the situation for the southern corn leaf blight fungus. After examining collections of H. maydis, it was determined that race T was present in many parts of the world some 7 to 15 years before the 1970 epidemic. However, the fungus existed mainly on gramineous hosts and not on corn because commonly planted varieties of corn were resistant to this race. Therefore, corn breeders could not have predicted the need to incorporate race-T resistance into their new corn lines. The second factor to consider is the extensive use of hybrid corn containing the Tams genetic background. In the 1930s, breeders began to capitalize on the phenomenon of hybrid vigor. When two inbred lines are crossed or hybridized, the resulting seed corn will produce a crop with enhanced agronomic traits, including enhanced yield. To accomplish these crosses efficiently in corn, breeders must remove the
31 mate flowers from the female plant to prevent sel£pollination. This was classically done by hand or machine. However, in the 1950s, cytoplasm~c male sterility was discovered and incorporated in corn breeding programs. By 1965, nearly 80 percent of the entire U.S. corn crop was produced with maTe-sterile techniques, specifically the use of Tcms What the breeders did not know, however, was that hybrid corn with a Tams genetic background was very susceptible to race T of N. maydis. ~ 1970, with proper weather conditions for disease development, with 85 percent of the corn crop containing Tams' and with an abundant supply of race T ~noculum, a southern corn leaf blight epidemic developed. Fortunately, however, the genetic basis for race-T susceptibility was quickly determined. By the next growing season, enough non-Tcm~ seed was available to farmers that losses were minimized. Evidence for the molecular basis of Tams activity has been oW tanned. Forde and Leaver (1989) reported that a polypeptide of 13,000 relative molecular mass (Mr) was unique to Tams mutochon- dria and that its expression depended on the activity of a nuclear restorer gene (a gene that overcomes the effect of cytoplasrn~c steril- ity). Dewey et al. (1987) identified the m~tochondrial gene encoding the 13~000 Mr polypeptide and deterrn~ned that the protein was am sociated with the rn~tochondrial membrane. Rottmann et al. (1987) demonstrated that male sterile TCm2' plants that mutated to mate fertile plants lost their ability to produce the 13,000 Mr polypeptide and that the mutation occurred in the area of the m~tochondrial genome that contains the gene for the 13,000 Mr polypeptide. ~ an effort to determine whether this polypeptide was also connected to increased susceptibility to H. maydis, Dewey et al. (1988) transferred the gene to Escherichia cold and demonstrated that bacteria produc- ing this polypeptide were sensitive to H. maydis toxin. Therefore, the gene for the 13,000 Mr polypeptide may have a pleiotropic effect in that it confers both male sterility and susceptibility to H. maydis. The story of Tams is given here to illustrate the types of potential problems that have developed ~ a result of the introduction of new variants. The southern corn leaf blight epidemic was a highly publicized event: an epidern~c ensued, and economic loss resulted. The year 1970 was certainly a bad year for corn production, but it was by no means a national catastrophe; corn production was back to aIrnost normal within a year. Because an occasional unexpected crop Toss may occur, it is unportant to have an arsenal of genetic
32 modification techniques and genetic resources available that can be used promptly to limit unacceptable losses. New molecular methods for gene introduction will be beneficial in this regard. Steady progress in the refinement of corn tissue culture sys- tems (Vasil, 1988), coupled with the development of electroporation (Frorntn et al., 1986) and particle-gun technologies (Klein et al., 1988), suggest that successful corn transformation may be imminent. ~ansgenic corn plants have been produced (Rhodes et al., 1988~; al- though these plants were sterile, this accomplishment demonstrates that significant progress is being made to develop gene transfer sys- tems for this important crop. PAST EXPERIENCE WITH CONFINEMENT Confinement is defined as any system of growing plants In which contact with plants of the same type is rn~nimized or plants are kept in defined areas. Plant breeders traditionally use confinement proce- dures to minimize genetic contamination of their field plots by pollen from outside sources such as neighboring fields. In addition, confine- ment practices are used to keep plant pathogens from spreading into or out of experimental field plots. Agricultural research, therefore, has a long history of experimentation that has been confined or kept within bounds. Both the private and the public sector, notably the land-grant in- stitutions or the Agricultural Research Service of the United States Department of Agriculture (USDA), undertake the first of several stages of cultivar development. For example, cultivated varieties of wheat are the result of 7 to 14 years of research and testing by both the public and private sector before marketing (Table 3-1~. Dur- ing this time, small numbers of plants are grown at selected sites and kept under close observation for environmental or organismal effects on the plant. Extensive records are usually compiled and, in the public sector, summarized and published. Few lines (or poten- tial varieties) survive such rigorous testing. Even after commercial use in farmers' fields, the plant's performance Is examined periodi- cally, by both sellers and producers of the seed or other propagative material. Some extremely well-adapted and highly productive culti- vars have a long commercial life, because of desirable characteristics that are difficult to improve. Other cultivars survive only a short time, perhaps five years, before they are replaced by higher yielding, disease-resistant, or otherwise improved cultivars. Biotechnology has
33 TABLE 3-1 An Illustrative Wheat Breeding Program Year Generation Activity Area (acres) 1 Make 300 to 400 crosses between varieties 0.1 or germplasm materials. 2 F1 Grow in field, greenhouses, or both 0.1 3 F2 Grow as bulk hybrid, evaluate for agronomic 0.5 and disease traits; quality evaluations for milling, mixing cunres, and protein content 4 F3 Bulk seed select determined number of heads 1.O from best crosses 5 F4 Head row nursery; 50,000 to 60,000 entries, 4.0 screen for disease resistance, select 5% on basis of resistance and plant type 6 F5 Preliminary observation nursery; agronomic 2.0 value; disease resistance; quality evaluations a for milling, mincing curves, and protein content 7 F. ~ Duplicate plots at one or more locations 2.3 8 F67a Preliminary yield trials at several locations 1.5 9 Ma Intrastate yield nursery at several locations 1.7S 10 Fga Station plots, on-farm tests, regional 4.0 . . ~ nurseries, merease seen 11 F10-13 Repeat testing; large-scale milling and 30.0 baking evaluations; seed Increase; name and release to certified growers . Equality evaluations for milling, baking, mincing properties, and protein content. the promise to shorten the cycle to commercial availability by two or more years through specific gene transformation and identification of the particular genes conveying desirable attributes. Tm the multiyear process of development of a useful cultivar, it is crucial to confine the seed and plants to the appropriate sites and to maintain the identity (purity) of the material (Table ~2~. This is done by confinement practices, which limit the plants or their products to a particular site and also protect neighboring fields from contaminating pollen. In this way, any unexpected effects can be observed. The distances cited ~ Table 3-2 are not absolute, but allow for acceptable levels of contamination. Specific information about the environment in which a cultivar was developed is necessary to make helpful site recommendations about suitable cultivars. Confinement as practiced by plant breeders or plant patholo- gists may be achieved in several ways. Simple confinement may be accomplished by the choice of an isolated location. Border rows for
34 TABLE 3-2 Isolation Requirements for Production of Genetically Pure Seed for Certain Species of Field and Vegetable Crops Type of Pollination Species Self- pollinated Self- pollinated but to a lesser degree than those listed above Cross Barley, oats, wheat, rice, soybean, lespedeza, field pea, garden bean, cowpea, flax grasses (self-pollina ted and apomictic species) Cotton (upland type) Cotton (Egyptian type) Pepper Tomato Tobacco Alfalfa, birdsfoot trefoil, pollinated by red clover, white clover, insects sweet clover Millet Onion Watermelon Cross- Hybrid field corn pollinated by wind Grasses 900 feet Isolation Distance for Highest Level of Genetic Purity Fields should be separated by a definite boundary adequate to prevent mechanical mixture 60 feet 100 feet from cultivars that differ markedly 1320 feet 200 feet 200 feet 150 feet or by four border rows of each culti~rar. Isolation between culti~rars of different types should be 1320 feet 600 feet 900 feet 1320 feet 5280 feet 2640 feet 660 feet (may be reduced if field is surrounded by specified numbers of border rows and the culti~rars nearby are of same color and texture) ADAPTED FROM: Association of Official Seed Certification Agencies, 1971. plants win limit both entry and exit of Insects or diseases that might otherwise harm the plants of interest. Fencing limits animal access. In tests conducted on a small scale, one uses the smallest numbers of plants that will give the information desired. More elaborate bar- riers to limit dispersal beyond the site include removing pollinating organs from plants, bagging flowers, and adjusting the time of year the plants are grown to avoid insect pests. Multiple physical and biological barriers are used in research plots and often In co~nmercial agriculture as well. Such barriers also include darns, soil terraces,
35 TABLE 3-3 Time Frames and Methods for Mitigating Unwanted Effects of Plants Immediate Short-term Long-term (Hours to several days) (0 to 3 yeare) (More than 3 yeare) Burning (eradication) Quarantine Tillage c Chemicals- Biological control Imgation/flooding Insect vector control Machinery sanitation Runoff water control Solarization Scourer with plastic) Breeding for resistancea Biological control- Quarantine Chemicals Crop rotation Culti~rar rotation Irrigat ion /flooding Heat treatment Soil solarization Induced resistance Meristem/tissue culture Insect Rector control Weed control Erosion control Breeding for resistance Biological control Crop rotation Cultivar rotation Soil amendments Weed control Erosion control Gerrnplasm may be adequately identified for rapid development; otherwise the process normally takes 5 to 10 years. brew biological control agents are yet available for widespread use; several are under investigation and development for some disease-causing microorganisms. Choice and availability of chemical for target plant and associated microorganisms dictate feasibility and approach. ADAPTED FROM: A. K. Vida~rer and G. Stotzky, 1989. tiliage practices, and the use of cherrucal or biological agents for con- tro} of insects or fungi. If necessary, physical barriers and security against unauthorized persons may be needed Biological barriers include genetic modifications to produce ste- rility or to recluce the ability of the plant to survive or escape preda- tors. The removal of reproductive organs and the removal of organ- isms that are hosts for a pathogen or insect can also be used. Death (normal decay), plowing under, and incineration are possible. Collectively, these procedures work well in research and usu- aDy very well in commercial use to protect human headth and the environment. If these common practices lose effectiveness, various ways of rrutigating deleterious effects are available (Table 3-3~. Some of these means are inexpensive and can be applied quickly, while others may
36 be costly and require longer periods to be effective. All these methods are applicable to genetically modified plants. SUMMARY POINTS 1. Techniques of genetic modification of plants were divided into three broad categories for the purposes of this report: cIassi- cal, cellular, and molecular. These techniques offer a wide array of possible genetic modification. Classical techniques mclude breed- ing by sexual hybridization, embryo rescue, undirected mutagenesis, and anther and ovule culture. Cellular techniques include cell fu- sion and somaclonal variation produced by tissue culture. Molecular techniques include directly introducing genes by a variety of trance formation procedures. 2. The results of genetic modification of plants are usually divided into two categories: those that increase yield and those that increase reliability of performance. Although these modifications can affect the persistence of plants, it will be Circuit to increase overaD persistence of domesticated crops because may persistence-related traits have been eliminated through breeding. 3. Plant breeders have a long history of safe field testing and introduction of many genetically modified crops. When problems occur they have been manageable and for the most part confined to the managed ecosystem. 4. Routinely used methods of plant confinement offer a vari- ety of options for limiting both gene transfer by pollen and direct escape of the genetically modified plant. Methods of confinement include biological, chemical, physical, geographical, environmental, and temporal control as weD as lirrutation of the size of the field plot.