Popular Science Monthly/Volume 77/October 1910/The Role of Hybridization in Plant Breeding
|THE RÔLE OF HYBRIDIZATION IN PLANT BREEDING|
By Professor E. M. EAST
THE word hybridization has been used with many meanings. The term is used here to denote the crossing of any two plants that differ from each other in a heritable character, whether they are of the same or of different species.
There is intimate connection between the rôle of hybridization and the rôle played by selection. It comes about in this way. Inherited variations are produced by nature with considerable profusion. New characters appear and old characters are lost: these form the working basis of selection. But whether they are large or small they are usually inherited completely. They are the units of heredity; or, if they are sometimes transmitted in units of lesser degree, they may be compared to chemical radicals.
The main object of hybridization then is the shuffling of these units in the first hybrid generation and their recombination in the next generation. There are, however, various phenomena attending hybridization, and I will endeavor to illustrate the following as those of most importance: (a) Recombination of characters and their fixation, (b) production of desirable combinations in the first hybrid generation and their continuation by asexual propagation, (c) production of fixed first generation hybrids, (d) production of blends.
If we begin at the real beginning in this discussion, we must say a few words concerning the actual mechanical operations of crossing. The first foundation stone to be laid is a knowledge of the flowering habits and flower structure of the plants to be used. Of course a careful examination of the flowers will show the easiest and surest method of removing the stamens of the flowers that are to be pollinated and of protecting them from foreign pollen. What is not so easily determined are the precise conditions under which the cross should be made to be successful. The proper preparation of the breeding plot even before the plants are grown is necessary. One takes it for granted that some fertilizer will be used, for the plants must be normal to seed well. The three essential elements of soil fertility are nitrogen, potassium and phosphorus, and to get the best results compounds of these elements must be present in proper proportions. First, available potash must be present in normal quantity although a certain excess will not be harmful. If nitrates are present in excess, however, vegetative growth will be over stimulated and seed production will be small. A lack of phosphates will produce the same effect upon seed production, but for a different reason. Phosphorus is an essential constituent of the proteid compounds found in large quantities in the seed. Therefore, if the plants are to be in the best condition for crossing, the soil should contain just the right amount of nitrates for normal growth and a generous supply of potash and phosphates. The exact amounts must be determined by experience for each soil and each species of plant.
Other necessary knowledge that can be obtained only from experience is which are the best flowers on the plant to serve as parents of the cross and what is the proper time for their pollination. For example, in the grasses the first flowers that appear usually form larger, healthier seed than the later blossoms. In most of the Solanaceæ, the petunias, browallias, etc., the exact opposite is true. The time when the individual flower is most receptive to pollen is even more narrowly limited. Both premature and delayed pollination is the cause of many failures and the optimum time should be accurately determined. Having exercised these precautions, it remains to study carefully the structure of the flower in order that it may be emasculated—i. e., the anthers removed before the pollen is shed—with sufficient adroitness that neither the anthers shall be opened nor the parts of the pistil injured. Only a few buds upon a single flower spike should be operated upon if they are to be given the best chance of development. If the buds are very small and some pollen unavoidably reaches them, it may be washed off with comparative safety with a dental syringe if done immediately. It is often recommended that the calyx and corolla be cut away when emasculating. This should be avoided if possible and the floral envelopes left as a protection to the pistil. After emasculation the buds should be protected from foreign pollen until time for pollination, and again after pollination at least until the fruits have begun to form. This protection may be an ordinary paper bag when the crossing is done in the field. In the greenhouse I find that a square of thin celluloid rolled around the flower and caught with two rubber bands, each end being protected with absorbent cotton plugs, is a better device. It gives excellent protection and allows transpiration.
But enough of the technique of hybridizing; the phenomena attending it are of more importance. After the pollen is placed upon the stigma it begins to grow until it reaches the ovule. Down this tube comes the male cell which contains the potentialities of its parent plant. This cell fuses with the female cell in the ovule and fertilization is accomplished. From this combination the seed and later the hybrid plant results, half of its characters coming from the plant which furnished the pollen and half from the plant to which it has been applied. If one studies the characters of several such hybrids, he finds many surprising facts. It usually makes no difference which plant is the mother plant, the result is the same. Certain characters are found in the hybrid that are identical with those possessed by the male parent and other characters the same as those possessed by the female parent. Other characters appear to have resulted from the blending of those of the two parents, while still others appear to be entirely new. The plant may be sterile if the cross is between widely differing species, but if it is fertile and the flower of the hybrid is self-fertilized, the plants resulting from this seed present still more surprises. For example, if one has crossed a pear-shaped yellow tomato with a round red tomato, in the second hybrid generation he will find individual plants bearing fruit of four kinds, pear-shaped yellow and round red, as were the two parents, and in addition pear-shaped red and round yellow. In other words all possible combinations occur and in definite proportions. Stated as a principle it may be said that where either of the parent plants possesses characters absent from the other, the potential characters remain pure in the germ cells of the hybrid and recombine as if by chance. This is the most important feature of the only law of heredity of which there is any exact knowledge—the law of Mendel. Let us illustrate the action of the law. Such a character as starchiness, as shown in "flint" maize, is either present or it is not present. The flinty appearance of the seed is due to the possession of some character that causes the maturation of plump starch grains. When this character is absent, the seeds dry up without maturing their starch grains, and present the wrinkled appearance common to sweet maize. Pairs of characters such as these, affecting a certain plant structure, are called contrasted or allelomorphic pairs. When a sweet maize is crossed with a flint maize, the resulting seeds are all flint like. That is, the dominant character or the character that calls for the presence of the structure or compound in question, manifests itself in the first hybrid generation. Complete dominance, however, is not a general phenomenon in crosses and as its importance is slight as compared with the second law, that of segregation of the pure characters (potentially) in the germ cells of the hybrid, we will discuss it no further. The second law predicts that in the generations succeeding a cross, plants grown from the selffertilized seeds of the hybrid reproduce both contrasted characters in the proportion of three of the dominant or "presence" characters to one of the recessive or absent character. Furthermore, inbred or self-fertilized plants bearing the recessive character continue ever after to breed true, while of those plants bearing the dominant character one third are pure and breed true while two thirds are hybrids and again throw the recessive character in one fourth of their offspring.
Fig. 1. Castration of Buds of Nicotiana tabacum before Crossing, a, correct stage for castration; b, method of slitting the corolla; c, castrated bud.
The theory supposes that when a dominant and a recessive character meet in a cross, the germ cells which are produced in the hybrid do not blend these characters, but possess either the one or the other; and as the possession of either character is a matter of chance, on the average 50 per cent, will bear the dominant and 50 per cent, will bear the recessive character. In a plant, for example, 50 per cent, of the pollen cells would bear the dominant and the other 50 per cent, would
Fig. 2. Castrated Bud of Impatiens sultani showing Method of Protection in Greenhouse Work.
bear the recessive character. One half of the egg cells, likewise, contain the dominant, and one half the recessive character.
Now, if we could pick out at random any one hundred pollen or male cells to fertilize any one hundred egg or female cells, we can see that there are equal chances for four results. A dominant male cell might meet a dominant female cell, a dominant male cell a recessive female cell, a recessive male cell a dominant female cell, and a recessive male cell a recessive female cell.
We have (D + D), (D + R), (R + D), and (R + R) plants formed in equal quantities, but as the two middle terms are the same, we can reduce the formula to one (D + D) to two (D + R) to one (R + R). But wherever there is a D present in the germ cell, the
Fig. 3. The Course of the Pollen-tube in a Rock-rose (Helianthemum marifolium). After Kerner and Oliver. 1, single flower; 2, essential organs of flower: course of pollen tubes shown diagrammatically; 3, pollen tubes penetrating the tissue of the pistil; 4, dried pollen grain; 5, pollen grain germinating; 6, ovule.
dominant character shows, while the recessive character is hidden. The one part or 25 per cent, of the individuals showing the character (D + D) will appear just like the two parts or 50 per cent, of the individuals having the character (D + R). Therefore, there will be 75 per cent, of the individuals which will show the dominant or D character, while 25 per cent, will show the recessive or R character. These 25 per cent, showing the R character will ever after breed true, because they contain nothing but the recessive character; while of the 75 per cent, showing the dominant character, one third or those having the pure (D + D) character will breed true in succeeding generations, while the other two thirds having the (D + R) or hybrid character will again split in the next generation.
For all practical purposes in plant breeding the mere fact of segregation is of greatest importance and the complexity of recent Mendelian interpretations need not bother us. Suffice it to say that most plant breeders have accepted the explanation that the recessive character is simply the lack or absence of the character in question, while the dominant character is its presence. This is simply a slightly different interpretation of the same facts and simplifies some of the more complex results of crossing. Instead of 50 per cent, of the germ cells bearing the flint character and 50 per cent, bearing the sweet character when sweet corn is crossed with flint corn, one should think of all of the germ cells bearing the ability to produce the wrinkled sweet corn seeds, but that 50 per cent, of them contain in addition the presence of a flint or starch producing character. In other words, the "starchy" character is superimposed upon the "sweet" character. The dominant and recessive characters in such a cross, then, are simply the presence and absence of the starchy character.
When several character pairs differentiate the two parent plants in a cross, all possible recombinations are formed, the relative frequency with which the combinations occur being simply the algebraic product of as many of the simple ratios as there are character pairs.
The importance of these Mendelian facts to the commercial plant breeder is great. In crossing plants differing in several simple characters that segregate after hybridization he may rest assured of two things. First, that with a sufficient number of progeny in the second hybrid generation, every possible recombination of the characters present will be represented by at least one pure specimen. Second, that these pure specimens when selfed, or pollinated with their own pollen, will breed true. It should be remembered, however, that one may have to self a number of plants to get the combination desired with all characters pure, for if any dominant characters are concerned, their purity can be ascertained only by breeding for another generation. As an illustration we may take the snap dragon, Antirrhinum. There is a long series of colors that segregate. There is also a type called the "Delilah," where the tube of the corolla is uncolored. Starting with this form in only one color, the whole color series of Delilah forms may be reproduced by crossing with the self colored strains. Or, one may combine the dwarf habit of growth of the Dwarf Champion tomato, with any of the various colors and shapes now on the market which have the ordinary tall habit of growth. Sometimes a very simple recombination is of very great commercial value. The so-called Havana type of wrapper tobacco grown in the Connecticut River valley has large leaves and a short stocky habit of growth. It produces from nineteen to twenty-one leaves. There is another type grown under cheese cloth shade which has a tall habit of growth with about twenty-six smaller leaves. The tall slender habit of growth makes it an undesirable type to grow in open fields where it is apt to be blown down. Mr.
Fig. 4. Showing Absolute Segregation in Second Hybrid Generation. These red and white ears grew from a single self-pollinated ear of the first hybrid generation of a cross between red and white maize.
A. D. Shamel, of the United States Department of Agriculture, crossed these two types. A new type called the Halladay has been produced with the higher number of leaves of the Cuban parent and the stocky habit of growth and large leaves of the Havana parent. The first interpretation of this result was that an entirely new variation had appeared, for the Cuban type usually has but twenty-two or twenty-three leaves. The writer has been able to show, however, that the actual strain of the Cuban used as the parent of the cross has on the average twenty-six leaves, and data have now been collected that show that the new variety is a simple recombination of the characters possessed by the two parents giving an out-door type averaging thirty per cent, greater yield than the old Havana strain. In a similar way Biffen has produced a rust resistant high-yielding wheat by crossing two varieties each of which possessed but one of these desirable qualities. Orton has combined the edible quality of the watermelon with the wilt resistance of the citron, and Webber has increased the ability of the orange to resist cold by crossing with the hardy trifoliate orange.
Recent accurately controlled investigations in hybridization have shown that many apparently complex results yield to simple explanations
Fig. 5. Red Maize Ear with Pericarp removed, showing segregation of yellow and white endosperm beneath it.
by use of the Mendelian theory. For example, two or more hereditary factors may be necessary for the production of an actual tangible character. If factors A and B must be present for its production, then a plant carrying only factor A and another carrying only
Fig. 6. Mendelian Segregation in Maize, a and b, the two parents, starchy and sweet maize; c, the first hybrid generation showing dominance of starchiness; d, the second hybrid generation showing segregation with the ratio of three starchy to one wrinkled seed. Lower row daughters of d. e, f and g, results of planting starchy seeds. One ear out of three is pure starchy, h, result of planting sweet seeds. Ear is pure sweet.
factor B do not possess the character. But let the two plants be crossed and the character appears. There are two white varieties of sweet peas; each, however, contains one of the two factors necessary for the production
of a purple variety. When these two white varieties are crossed, the purple variety results. The second generation, however, produces seven whites to every nine purples. Such segregation into purples and whites may not he desirable; all purples may be wanted. This brings us to a consideration of class B of the four classes of phenomena attending hybridization, the production of desirable character combinations in the first hybrid generation and their continuation by asexual propagation. This class really includes several distinct types of occurrences. The purple sweet pea produced from the two whites will serve as an illustration of the first type. In certain plants (not meaning the sweet pea, however) it is as simple to reproduce by cuttings as by seed. The cuttings are simply parts of the plant from which they come and are identical with it in character. If in a species of this kind a desirable character is formed by the union of two or more hereditary factors and one wishes to reproduce the character indefinitely, asexual reproduction by cutting serves the purpose admirably.
There is another case of a different kind. Sometimes the hybrid character is different from the character of the parents, even though the exact parental characters are reproduced by segregation in succeeding generations. The commercial carnation form is the result of crossing the single carnation with the huge worthless doubles called "busters." Reproduced by seed the commercial carnation throws both singles and busters, showing that segregation of the parental characters takes place; but as these plants are easily reproduced by cuttings, and the cuttings are all of the commercial type, sexual reproduction is only resorted to for the sake of producing new varieties. Another common phenomenon attending hybridization is sterility. Many very beautiful flowers produce no seed at all. This is even an advantage in some cases, because the plants flower more profusely than if they were spending their energies in the production of seed. Here again, cuttings are resorted to to reproduce the hybrid, or, as in the case of seedless oranges, the cuttings are grafted into an older rootstock instead of being rooted.
I stated at the beginning that there were two other classes of hybridization phenomena, the production of fixed first-generation hybrids and the production of blend hybrids. It is probable in the last analysis that the true explanation of these cases is the same; so we will consider them together. It is believed by many that there are kinds of inheritance other than Mendelian, that is, inheritance where no segregation occurs. Far be it from me to deny this; I simply state the fact that there are no exact data extant proving other kinds of inheritance. Such data may be found, but it is useless to speculate upon other laws without such evidence. There are several cases in which either new characters that breed true or blended characters that breed true appear to have been formed, but they have not been studied with sufficient care for an analysis of their mode of inheritance to be accurate and final. It is in crosses between true species that hybrids have been formed seemingly as constant and uniform as their parent species. Janczewsky has produced several such hybrids. Perhaps the most famous, however, are the blackberry-raspberry crosses first produced by the late E. S. Carman, editor of the Rural New-Yorker and later by Luther Burbank and others. Several hybrids having a commercial value have been made in this genus (Rubus), and all of them reproduce approximately true from seed. These are the facts and show what may sometimes be expected by hybridizers when crossing true species; but I wish to point out that this does not necessarily mean that we are dealing with a new mode of inheritance. Bramble species produce seedlings that are quite variable and in which the variations are extremely difficult to describe; there is, therefore, no exact information as to the relative variability of the hybrid seedlings as compared to that of the two parents. It may be said, then, that it is yet unknown whether there is partial segregation.
Fig. 8. First hybrid Generation of Cross between "Havana" and "Cuban" Varieties of Tobacco shown in Fig. 7. Plant is taller than either parent showing the increased vigor due to a cross. Size of leaf of "Havana" is dominant. Habit of growth of "Cuban" is dominant. Number of leaves is intermediate, but approaches the "Cuban."
But why should there not be complete segregation to the types of each parent? In the first place, because it is likely that numerous separately heritable characters are concerned, and when n pairs of characters are concerned it takes four to the nth power seedlings to run an even chance that there will be one plant like each of the parents. When we consider that with ten pairs of characters, this means over 1,000,000 individuals, we can see with what enormous numbers one has to deal. In the second place these hybrids are only partially fertile, and as I have suggested in former papers, some consideration must be given the fact that there may be selective fertilization that works against extreme
Fig. 9. Recombination of characters of Plants shown in Fig. 7, occurring in the second hybrid generation. This is a uniform and constant type having the short habit of growth and large leaves of the "Havana" parent, combined with the high number of leaves of the "Cuban" parent. It is now grown in the Connecticut River valley and yields 40 per cent, more than the Havana type.
segregation. To take a hypothetical case, suppose two plants are crossed in which the flowers of one are twice as long as the flowers of the other and that this extra length is controlled by three or four separately heritable factors. If only a few of the egg cells can be fertilized on account of dissimilarity from the pollen cells, one would expect only those seeds to be formed that would come from the fusion of the germ cells nearest alike. Intermediates would therefore be more likely to be formed than extremes. There is one other possible way of accounting
for constant intermediate hybrid races which I think has never before been mentioned. In crossing species of the genus Nicotiana, I have had plants develop from seed that have apparently been formed apogamously, that is, formed from an immature egg cell without fertilization. It is evident that this is induced by the extraordinary irritation of foreign pollen. The true hybrid plants that are formed are generally blends in the first generation. The question, then, arises: May not the. difficulty of maturing sexual cells in a wide cross sometimes cause apogamous development and therefore a continued propagation of a constant and uniform race?
All but the last of these suggestions may also be pertinent in the case of varietal crosses where there is said to be a blending of characters that deal with size. I am not certain, however, that all the so-called blend hybrids might not show segregation if studied in large numbers. I have found such segregation in size characters in crosses of both maize varieties and of tobacco varieties.
- This paper is based on one of a series of popular lectures delivered at the Bussey Institution of Harvard University, April and May, 1910.
- There are certain cases like variegation that are exceptions to this rule.
- In the writer's paper "The Role of Selection in Plant Breeding" in the August number of this journal, the legends for figures three and four unfortunately were interchanged in printing.