PRACTICAL METHODS OF SELECTION WORK
BASED ON THE THEORY OF DEVELOPMENT

The Institute of Selection and Genetics (Odessa) has resolutely proceeded to put the theoretical premises here briefly enunciated into practice to reorganize the work of genetic breeding from the standpoint of the theory of development.

In this reorganization we strive to take the Marxist-Leninist theory of development, which is the only true theory, as our guide. It is our aim to apply this theory specifically to the problem that we are working on.

All this has enabled us now not only to conduct extensive work in controlling the individual development of plants, but also, in theory and in practice, to trace a new path for plant-breeding work, to take the course of vanquishing formalism and empiricism in genetic-breeding research.

Table 3

EARlNG F2 1934

Combination 25/V 26 27 28 29 30 31 1/VI  2 3 4 5 Total
earing
earlier
than
Lutesc.
062
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Did
not
ear
Total
plants
earing
later
than
Lutesc.
062
Total
eared
Ersp. 534/1 x 0274 Girka - - - - - - 4 11 3 33 5 120 176 31 172 116 35 9 12 3 71 3 34 3 14 4 12 4 - 2 5 - 1 1 1 - 2 - 44 579 711
Ersp. 1637/1 x 062 Lutesc. - - - - 1 - 1 5 6 1 4 16 34 - 9 72 8 3 5 12 9 3 - - - - - - - - - - - - - - - - 29 142 147
Ersp. 1637/1 x 0274 Girka - - - - 2 - 27 23 51 35 10 197 345 17 403 104 55 35 28 11 92 18 52 - 2 - 9 - - - 1 - 1 - 2 - 2 - 186 1021 1180
Ersp. 2522/1 x 062 Lutesc. - - - - - - 6 25 34 26 12 128 231 - 36 128 34 88 18 6 53 8 3 3 - - 2 1 - - - - 1 1 1 - - - 26 409 614
Ersp. 2522/1 x 0274 Girka - - - - 1 1 4 15 44 12 - 123 199 15 121 37 34 5 17 7 49 3 2 5 - - 5 - - - 2 - 4 - - - - - 24 330 505
Ersp. 3030/15 x 062 Lutesc. 1 - - 3 2 - 25 93 81 17 7 202 431 1 27 232 38 50 22 12 87 50 5 22 9 - 2 3 - 2 1 - 5 1 - - - - 57 626 1000
Ersp. 3060/15 x 0274 Girka - - - - - - 9 15 11 57 1 97 190 183 224 43 32 35 8 6 57 65 11 11 18 1 7 3 1 1 5 3 4 1 4 2 - 1 28 757 919
Ferrug.1316/2 x 062 Lutesc. - - - - - - - - 1 - 1 - 2 5 10 8 - 2 - - 7 - 2 - - - - - - - - - - - - - - - 2 37 37
Total   1608   3901 5113

[97]

These premises have enabled us to indicate exactly and according to plan how long it will take to raise varieties of spring wheat for the Odessa district. These premises mark a considerable step forward in overcoming the element of chance in breeding work, for they make it possible to plan the output of varieties with ever-improving indices according to schedule. The approach to the hereditary foundation from the standpoint of development creates real possibilities of making the entire work of plant breeding a much more intelligent process because, if one bases himself on the laws already ascertained one can keep constant check on the work and, in the course of it, note and rectify errors in that process.

After choosing the parental forms for crossing with the object of obtaining forms more early ripening than either parent, the correctness of this choice can be tested by the plants of the very first generation. If F1 is later ripening than intended for the future desired variety, it is a sure sign that a mistake has been made in the choice of parental forms. On the assumption that the segregates of F2 and of succeeding generations cannot be earlier ripening than F1, it is possible in F1 to cull the combinations with respect to length of vegetative period.

Thus, no variety of spring wheat that ripens later than Lutescens 062 from the Saratov station is suitable for Odessa conditions. This, of course, does not mean that any early-ripening variety will be suitable here.

Hence, when it proceeded to breed a spring variety for Odessa conditions, the Institute of Selection and Genetics resolved beforehand that the future variety of spring wheat must under no circumstances ripen later than Lutescens 062. It must ripen at least 3-4 days earlier.

[98]

This line of action makes it possible to perform the first culling of the breeding material in the first generation-something which plant breeders usually did not do before. If there are several combinations of F1 plants in the field, all the combinations which in F1 are later ripening than Lutescens 062, or have the same length of vegetative period, must be culled. Only those combinations must be left which in F1 are earlier-ripening than 062.

We will remind the reader that such culling has become possible only because of the law that we have discovered, namely, that not a single segregate in all succeeding generations can be earlier ripening than the initial heterozygous forms of the preceding generation, and cannot be earlier ripening than F1. By culling combinations in F1, the plant breeder can, to some extent, save himself work and bother with combinations which obviously will not give him segregates with the necessary length of vegetative period, as he will definitely know beforehand that from such combinations he will not get a variety suitable for his district, in this case, Odessa.

After sowing F2, which will segregate in numerous characters, including length of vegetative period, as regards the course of the phases as well as the organs and characters that develop on the basis of the phases, it is necessary, first of all, to cull while still in the field all the segregates that ripen later than the variety it is intended to breed. Segregates which under the conditions of our district have the same ripening period as Lutescens 062, or are later ripening, will in any case be unable (in the absence of an inner basis for segregation) in succeeding generations to produce forms with the requisite vegetative period. After that it is necessary to proceed to the next breeding job, namely, to choose specimens from the remaining small percentage of segregates. In the breeding work conducted at the Institute of Selection and Genetics the following procedure has been adopted: in F2, only those plants are left which ripen at least two days earlier than 062.

Table 4

EARING F3 1934

Combination 29/V 30 31 1/VI  2  3  4  5 Total
earing
earlier
than
Lutesc.
062
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Did
not
ear
Total
plants
earing
later
than
Lutesc.
062
Total
eared
Ersp. 534/1 x 0274 Girka - - - 1 2 - 1 2 6 9 16 9 3 4 5 - 5 2 3 - 3 - 1 1 1 - - - - - - - 4 13 79 72
Ersp. 1637/1 x 062 Lutesc. 1 - 1 - 1 - - 2 5 4 6 8 1 3 1 - 5 2 - 1 - - - - 1 - - - - - - - 1 - 33 38
Ersp. 1637/1 x 0274 Girka - - - - - 2 - 6 8 13 12 51 9 11 17 1 11 3 2 - 15 - 1 - 1 - - - - - - - 1 11 160 157
Ersp. 2522/1 x 062 Lutesc. - - - - 1 - 1 1 3 11 5 9 5 2 9 - 1 1 - 1 1 - - - - - - - - - - 1 1 1 48 50
Ersp. 2522/1 x 0274 Girka 1 - 1 - 6 - 2 - 10 14 - 18 5 2 4 1 3 - - - 1 - - - 1 - - - - - - - - - 48 58
Ersp. 3060/15 x 062 Lutesc. - - - - 5 1 - - 6 5 2 4 5 - - - 1 - - - 2 - - - 1 - - - - - - - 1 - 21 27
Ersp. 3060/15 x 0274 Girka - - - - 5 - 2 4 11 25 7 27 14 5 8 - 6 1 - - 2 - 2 - 1 2 1 - 1 - 1 - 1 5 109 115
Ferrug. 1316/2 x 062 Lutesc. - - - - 1 - 1 1 3 1 - - - - - - - - - - - - - - - - - - - - - - - - 1 4
Total   52   499 521

[99]

For example, in the combination given in Table 3 Erythrospermum 534/1 x 0274 Girka of local breeding, out of 755 F2 plants, only 176 eared earlier than Lutescens 062 (plants enumerated on the left side of the table); the remaining 579 were culled. In the combination Erythrospermum 1637/1 x 062, out of 176 plants, only 34 were left. In the combination Erythrospermum 1637/1 x 0274 Girka, out of 1,366 F2 plants only 345 were left as regards length of vegetative period; the rest were culled when still in the field, etc. The small percentage of plants left after culling according to vegetative period is then subjected to the ordinary culling practised in plant breeding, which, of course, greatly reduces the number of remaining plants (degenerate, obviously unproductive, and other such forms are rejected), depending on the approach and on what is required of the variety that is being raised. The same is done in F3, F4, etc., until the constant requisite form for the conditions of the given district is obtained (Table 4).

Proceeding in this way, one may at once encounter the following obstacles: parental forms, properly chosen for crossing in regard to length of vegetative period, as confirmed by the first generation, may sometimes, in F2, the succeeding generation, fail to produce a single segregate, or give rise to exceedingly few that remain (in view of range of vegetative period) after culling in the manner described above. It becomes obvious to the plant breeder that the needed variety cannot be raised from the remaining forms, but at the same time the phasic analysis shows that the variety can be obtained from this parental pair. In such cases it is perfectly evident that such a combination must not be rejected. It merely indicates that in this case F2 was sown on too small a scale, and that this combination, the sowing of F2, must be repeated on a larger scale. The same will apply to all the succeeding generations.

Thus, the propositions we advance in aid of the practical work of the [100] plant breeder enable him to work all the time with ten and even a hundred times less plants than he has worked with up till now (or at least than were required by the genetic theory of hybridological analysis). At the same time he can now enlarge the scale of work tenfold, subjecting to selection combinations of all the existing varieties of the given crop, and greatly reducing the time required for raising varieties and the area of nurseries needed for this. Owing to the small number of new forms of potential varieties, each of them will receive ever so much more attention from the breeder than was the case before, under preliminary variety testing, when he had to deal with hundreds of claimants to be the future variety. It must be pointed out that choosing parental pairs for crossing according to phases of development, i.e., taking phases as one of the criteria in choosing parental pairs, does not mean that for crossing purposes those forms are chosen, or given preference, out of which a new form may be obtained phasically best suited to the conditions of the given district. If this course is followed it may safely be predicted [101] that any plant breeder will very easily breed at request only early ripeners. But as yet they will not be the varieties of, say, spring wheat needed for the steppes of the Ukrainian S.S.R., or for the Far North (where early ripeners are also needed). In practical work, it will become perfectly obvious to every plant breeder that in large collections analyzed according to developmental phases the choice of crossing material should not be confined to those forms that are phasically best suited for this purpose. Different forms may be phasically suitable for the conditions of a given district, but these same forms, with their excellent phases of development, may be unable to stand the dry conditions of the Odessa district, or other climatic and soil conditions, or may not meet the demands socialist farming in a given district makes upon the given crop. Therefore, when choosing parental forms by phasic analysis, it is necessary, before crossing, to ascertain the economic indices of these forms by phasic analysis under the conditions of the given district.

   
Fig. 34. Early-ripening F1 obtained from two late-ripening parents.
Choice of pair made on basis of phasic analysis
  Fig. 35. Pallidum 574/1 (Azerbaijan)
Late ripening owing to vernalization phase (sheaf on left). 943 late ripening owing to photo phase (sheaf on right). Phasic analysis showed that F1 should be of earlier habit than either parent, and it turned out to be so
  Fig. 36. Choice of sesame pair for producing an early-ripening form from two late-ripening forms.
Choice made on basis of phasic analysis

In the course of its work, the Laboratory of the Physiology of Plant Development (Institute of Selection and Genetics) crossed thirty pairs of wheat forms chosen according to phases of development. But before the crossing it was arranged that for practical purposes, i.e., for raising a variety for the Odessa district, the laboratory should cross only two pairs of forms: Erythrospermum 534/1 x 0274 Girka of local breeding, and Erythrospermum 543/1 x Lutescens 062 from the Saratov station. The bulk of the other crossings was made exclusively for the purpose of testing the possibility of combining new forms according to the length of the vegetative period, taking into account the phasic development of the parents; they were 'made for the purpose of testing the propositions put forward above. These crossings whose purpose was methodological, we, in jest, called "crossings for the camera," illustrations of which are given in this essay (Figs. 34-40). As is evident [102] from the illustrations, the pictures obtained were not bad. Every combination of crosses resulted in a form that was much earlier ripening than either parent.

Fig. 37. Choice of parental pair of barley for producing early-ripening form from two late-ripening forms.
Choice made by phasic analysis of hereditary foundation

We did not choose the first two pairs of combinations for breeding the variety because they were the best in the entire world collection. We are convinced that the collection contains better combinations of all possible kind's, but they have not yet been discovered. It is important for us, however, to ascertain the potentialities and developmental conditions of the hereditary foundation of the two components that are taken for crossing against the background of the length of the vegetative period of our future variety under the conditions of the

district for which it is being raised. As these conditions vary from year to year in each district, although they remain relatively constant (Odessa conditions in Odessa, Kharkov conditions in Kharkov, etc.), we must, before crossing, find out the economic indices of the parent forms under the district's conditions of growth over a number of years against the background of the vegetative period of the future variety.

If not vernalized, the Erythrospermum variety we took from the Azerbaijan station as the initial form for obtaining a variety of spring wheat for the Odessa district is, under the conditions of that district, an extremely late ripener and its yield is small. After vernalization, this variety, under Odessa conditions, produces a fairly good yield for three years. Girka 0274 of local breeding is less productive under Odessa conditions than Lutescens 062 from the Saratov station; and it ripens 5 days later than Lutescens 062. Both Lutescens 062 and Girka produce exceedingly low yields under Odessa conditions. A phasic analysis of these forms of wheat, however, shows that if their growth is accelerated (in the photophase) by 6-8 days, Lutescens 062 and Girka 0274 show good indices as regards fulness and quality of grains, i.e., the chief yield indices of spring wheat for our southern districts.

[103]

 
Fig. 38. Segregates of F3 are earlier ripening than either parent   Fig. 39. Segregates of F3. Earlier ripening than parents (middle sheaves)

In view of the fact that Girka is a local variety, there are more grounds for assuming that its hereditary foundation, if its "weak spot"—the photophase—is removed, will prove to be better adapted to the variations of the climatic conditions of the Odessa district. Lutescens 062 is not a local variety (it was raised at the Saratov station). But it was produced from Poltavka by individual selection; moreover, this variety has already been grown in the districts of Odessa Region for quite a number of years. This explains why we took these two combinations in hand in order to obtain quickly the variety we want. During the initial period, these two combinations are much safer insurance for us than other combinations, which, though more tempting as regards their phases of development, provide no guarantee as regards economic indices. The constant forms obtained from these combinations under Odessa field conditions ripened 10 days earlier than Girka (one of the parents) and 6 days earlier than Lutescens 062 from the Saratov station (the variety recommended for Odessa Region). We believe that these forms will be the varieties we have been seeking.

[104]

Fig. 40. Pair chosen on basis of phasic analysis to produce an early-ripening first generation from a winter and a late-ripening form

The plant-breeding work on the new lines which the Laboratory of the Physiology of Plant Development of the Institute of Selection and Genetics has been conducting since 1933, was begun with the growing of the parents. It has already resulted in constant forms and has been carried to F7. Beginning with the spring of 1935, the forms that turn out to be the best under the given conditions of development will be subjected to various field tests. At the same time, three or four forms, the indices of which we believe to be the best on theoretical grounds and which are being tested and verified in the course of different generations, will be reproduced, and by the autumn of 1935 each will be brought up to 10 c. of grain (thus passing in 2 years and 10 months from the parental forms to 10 c. of grain).

The proposed seemingly excessive forcing of the reproduction of a variety which has not gone through the usual preliminary varietal test is, in our opinion, not only possible but necessary. This is so because the theoretical propositions by which the Laboratory of the Physiology of Plant Development is guided in selecting parental pairs for crossing, and the theoretically assumed general pattern of segregation, have been confirmed by the whole course of segregation up to the appearance of the expected constant [105] forms and their characters. This imbues us with a certain amount of confidence that the field varietal tests will further confirm our assumptions concerning the varieties' economic qualities. These, we think, should be higher under conditions analogous to those of the district in which they were raised than the economic qualities of the standard varieties of the given district, not to speak of the original parental forms (Girka and Ersp. 534/1), which in many respects are inferior to the standard variety of spring wheat.

Fig. 41. Erythrosp. 2781 (Azerbaijan) ripens very late under Odessa conditions (pot on extreme left)
When vernalized shows good economic indices. Lutescens 062, late ripening under Odessa conditions (pot on extreme right). Under favourable conditions for passage through second (photo) phase becomes early ripening and shows good economic indices. When these are crossed a hereditary basis is obtained from which the sole "weak spots" have been eliminated. The cross should result in a variety with good economic indices, ascertained by means of the phasic analysis of each of these forms. Centre (two pots): F1 hybrids

The course and state of development of the hybrids on the experimental fields of the Institute of Selection and Genetics make us feel confident that the problem of breeding the varieties needed by our socialist agriculture will shortly be solved.

First published in 1935

BACK INDEX