Improvement of Wheat and Oats at Svalöf

by

H. NILSSON-EHLE.

Since 1900 the writer of this paper has been conducting the work of plant breeding in the Wheat and Oats Section at Svalöf; he has published a series of memoirs and notes, some theoretical, others practical, on the subject. In the "Beitrage zur Pflanzenzucht III" the writer gives a resumé of his work on winter wheat. A similar summary, in German, of the work accomplished on oats, with complete bibliography, is in course of preparation.

The present article proposes to give the most important results of all this work. The chief object aimed at during these twelve years has been to provide farmers with improved varieties. Secondarily, so far as time permitted, an endeavour has been made to develop further the principles and methods of practical breeding, by giving in a series of publications, partly of a purely theoretical nature, information on the bases of breeding, variation and heredity.

Practical breeding has resulted in a series of new varieties which have been handed over to practical farming, but only such varieties were distributed which in practical experiments yielded a heavier crop than those previously grown or otherwise revealed themselves superior.

These improved varieties were selected only according to yield and to direct valuation of their good qualities, and not according to an indirect opinion based upon a supposed correlation. An increase in the number of varieties grown in Sweden was vigorously avoided: the new varieties were so distributed as to replace the varieties previously grown, and no greater number of varieties were produced than those required by the very various climatic and other conditions obtaining in Sweden. The most important new varieties launched between 1900 and 1912 are: among winter wheats, Extra Squarehead II (1909) (1) and Sun wheat (1911) for the South of Sweder, and Poodle wheat for Central Sweden; among foats Conqueror (1908) for the white oat districts of the South of Sweden and Black Bell II (1909) for the black oat districts of Central Sweden. All these varieties spread widely during the succeeding years, not only in Sweden, but to a certain extent in neighbouring countries also, and have retained their position in practical farming. Besides the above, some few varieties adapted to more special conditions were distributed (Fyris oats and others).

(1) The date in brackets is that of the year in which the varieties were first placed upon the market and mostly in great quantities.

The new varieties have replaced the older ones: Extra-Squarehead II has completely replaced Extra-Squarehead I, the yield of which was considerably inferior (about 15 per cent); the Grenadier variety is also destined to disappear gradually. As for oats, the varieties Hvitling and White Probsteier, as well as Black Bell I formerly bred at Svalöf, have been withdrawn and they are no longer sold at Svalöf. Thus an increase in the number of varieties has not taken place. A few other varieties whrich show further improvement in certain directions have been given out to be grown on a large scale and will within a very few years be placed on the market; thus among oats, Crown and an early cross for the North of Sweden, and amorg wheats a very productive cross for South Sweden (0801) and Thule for Central Sweden. The multiplication of the most recent crosses, which represent the highest mark at present attained in improvement, is being carried out.

The above new varieties have been obtained partly by selection from native cereals and partly-during the last few years principally-by crossing. By mutation, on the other hand, no new improved variety has been obtained. The artificial crosses were made with the object of combining in the best manner the principal valuable qualities, such as resistance to cold, lodging and diseases, productivity, and the like, and at the same time of increasing these qualities in the desired direction. In the attainment of these aims continual progress has been made.

The work concerning the theoretical basis of breeding referred especially to the nature of the hereditary variation in cereals and to the manner in which several qualities, especially those of practical importance, are inherited. With this object the writer began in 1900, the year in which the great discoveries of Mendel were known, comprehensive experiments in crossing; the results have hitherto beer only partially published in the two principal works of the years 1909 and 1910, as well as in a series of shorter papers.

37 66

The first result of these researches agreed with those of Von Tschermak, Biffen, Spillman, De Vilmorin, Howard and others, namely that all the qualities of cereals, after crossing, show a sure segregation and reciprocal independence and form all kinds of combinations. The so-called "elementary species, constant forms, pure lines" of the mostly self-pollinating cereals are not, as De Vries thinks, independent unities obtained by progressive mutations, but various homozygotic combinations of segregating qualities. The crossing of two constant forms may give rise to an immense variety of new forms which may be maintained constant; it suffices thus to have only one such cross to render manifest all the diversity of constant forms or lines of a native breed. In other words, by the crossing of only two constant forms, a whole "population" of new forms and lines is obtained. As spontaneous crossing occurs frequently in self-pollinating cereals the nature of the so-called elementary species or lines as combinations of crosses becomes quite clear.

This fact being recognized, artificial crossing became naturally a specially important principle in breeding. The object of breding was no longer limited to the selection of the existing homozygotic combinations, but was exten

ded to the obtaining of always better combinations by appropriate crossing; and in this connection not only the crosses between older and more different varieties, but also between lines, partly of the same old breeds.

As for the magnitude of the differences between the homozygotic combinations ("genotypes" of Johannsen), it was shown in connection with researches on the variation of native breeds (Sveriges Utsädesförenings Tidskrift, 1901, pp. 154-176; 1908, pp. 165-170; Botan. Notiser, 1907, pp. 113-140), as well as upon the segregation of qualities after crossing, that in respect of most qualities, especially of the practically important ones, the genotypes are so numerous and so closely connected that they form fully continuous lines; the result is that by cultivating separately plants of similar aspect, an apparently uniform group of individuals can be split up into different descendants, into lines hereditarily different. Among the genotypes one may, it is true, isolate well characterised "elementary species" from a native breed, but a more minute analysis of the native breed reveals that they are corrected with each other by the finest transitions, and the differences are only in the rarest cases of a qualitative "botanical" nature, and this especially for those qualities with which practical breeding deals. Further, in quite similar morphological types the physiological characters most important in breeding, such as productivity, resistance to disease, lodging and cold, early maturity and the like, may be essentially different. In consequence of this, "separation by lines" instead of "separation by forms was introduced in the years 1901-03. Instead of separating out from old strains a relatively limited number of morphologically characterised forms and fixing them by continued selection, while further improving them by utilizing the new variations which appeared, as had been done until then at Svalöf also, further improvement in this direction was abandoned because no good results had been obtained (1); in its place a more extensive improvement of old varieties was resorted to, by selecting for separate cultivation from each variety to be bred, the greatest possible number, sometimes several hundreds of plants, without caring whether their outer appearance was the same or not, and their descendants were compared as to their production and practical value. From these descendants or lines the best were finally picked out. A further selection for the better fixation of the characters of the finally selected lines was not usually made, because with this separation of lines from old breeds of mostly selfpollinating cereals the lines in general were already sufficiently fixed. This is explained by the fact that by continued selfing only the homozygotic combinations can be maintained in the long run, as Mendel had already shown.

The descendance resulting from crossing was treated in a similar manner, only the separation of lines was frequently repeated. By this method of separation by lines all the new improved varieties have been bred.

(1) Further improvement was sought for by means of suitable crossing.

In the attempts made to obtain desired combinations from crosses, it must be noted in the first place that most of the practically important characters are very probably excedingly complex "construction characters" (Konstruktionseigenschaften). After crossing two constant forms or lines which represent two degrees of a character, e. g. size, resistance to rust or cold, time of maturing, the segregation is always complicated and can only be ascertained by the separate cultivation of all the F2 plants and by the comparison of the average characters of their offspring in F3. With regard, for instance, to the medium height of plants, the constant oat plants vary very much; there are taller and shorter lines in a whole series of degrees which could be represented by:

I 2 3 4 5 6 7 8 9 IO.

Crosses between any two tall and low constant lines do not give (or only in the rarest cases) the simple Mendeliar segregation. The segregation does not take place in the following manner:

Cross 1X8. F: 14 homozygote 1, 2 intermediate heterozygote, 4 homozygote 8.

1X5. F: 4 homozygote 1, 2 intermediate heterozygote, 4 homozygote 5.

5X8. F4 homozygote 5, 2 intermediate heterozygote, 4 homozygote 8.

On the contrary, the segregation after crossing for instance I×8 gives a whole series of gradations, among which those of the parents are very rare or may be completely wanting, so that the whole segregation is intermediate and does not attain the limits given by the parents:

Inversely, when two medium slightly different lines, such as 5 and 7, are crossed with each other, the segregation is very often, and sometimes eonsiderably, transgressive, that is degrees appear which exceed the limits of the parents in one or in both directions; lines are formed which on the average are taller or shorter than their parents and which on further cultivation show that they possess this character as a true hereditary trait:

As every hereditary line shows certain fluctuations in its modifications, which can sometimes be very considerable, the transgressive segregation might perhaps be better represented as follows:

The fact is also important that lines that possess the same degree of a character give on crossing segregations which form transgressive lines:

This is the manner in which all the quantitative practically important characters that the writer has investigated in wheat and oats behave; such, for instance, are the degree of resistance to cold (cf. Zeitschrift für Pflanzenzüchtung, Vol. 1, 1912, pp. 3-12), early maturity (cf. Compte rendu de la IV conférence intern. de génétique à Paris, 1911, pp. 136-155: Paris, 1913), resistance to rust and to lodging, productivity, size of grain. The extensive researches of Tedin on barley (short notice in Popular Naturvet. Revy, 1912, p. 216) show many quite similar results, in regard to early ripening, degree of emergence of the ear from the sheath, and so forth. Sometimes the transgressions after the crossing of constant lines of medium characters are very considerable (e. g. the time of ripening of oats).

On the basis of all the facts, agreeing for several characters, and concerning the behaviour of segregation not only in F2 but also in the following generation (in F3), it is to be noted among other points that the segregation in certain descendants is considerably greater than in others. The writer has suggested the theory that all practically important characters are of a complex nature ("Konstruktionseigenschaften ") determined by inner Mendelian factors, which through various combinations, form the long series of hereditary constant gradations.