Wide Hybridization in Plants (1960) p. 113-123
V. F. Lyubimova
Candidate of Biological Sciences
Main Botanic Gardens of the Academy of Sciences of the U. S. S. R.

A great difficulty often encountered when new plant forms and species are produced by wide hybridization is the sterility of first-generation hybrids and the reduced fertility of some later generations, among them the stable forms.

Sterility can be overcome and fertility of hybrids increased by various methods; by backcrossing in one or another form, by inducing polyploidy, by raising hybrid embryos on an artificial nutrient medium, by the utilization of microelements and physiologically active substances, and by the provision of optimum conditions during meiosis and the postmeiotic ripening of sexual cells when the hybrids flower and are fertilized.

The oldest and best known means of combating hybrid sterility is backcrossing to one of the parental forms; in such backcrosses the hybrid form is usually taken as the maternal plant. But the hybrid plant can also be utilized as pollen parent, if nondehiscing anthers are opened artificially, because normally developed and viable pollen grains are sometimes present among the bulk of abortive pollen. In some cases, first-generation sterility can be overcome by opening the coriaceous, nondehiscing anthers artificially and then pollinating the hybrid with its own pollen. By this method O. N. Sorokina succeeded in overcoming the sterility of the hybrid Aegilops longissima X Triticum durum var. apulicum, which did not set seed with other methods of backcrossing.

The overcoming of hybrid sterility by induced chromosome duplication is one of the most effective and well-known methods. It has gained wide recognition and has been used since the discovery by Blakeslee and Avery (1937) of the potent substance colchicine, which causes polyploidy. In some cases, induced polyploidy is the only means by which sterility can be overcome, particularly in hybrids with total asyndesis of chromosomes. By using colchicine, sterility has been overcome in many hybrids. Instances among the Gramineae are: rye-wheat hybrids (Pisarev, 1955, Sears, 1939); wheat-Agropyron hybrids (Armstrong and Lehman, 1944; Rawa, 1939; Peto and Boyes, 1941; Bao Ven' Kui and Yan' Yui-zhui, 1956); wheat-Aegilops hybrids (Sears, 1939); rye-Agropyron hybrids (Palamarchuk, 1948); interspecific wheat hybrids (Zhebrak, 1941; Kasparyan, 1940), and others.

In recent years, many papers have been published on overcoming incompatibility between hybrids, and on increasing hybrid fertility by using microelements and physiologically active substances, mainly in the form of dusts applied to the plants in their growing season.

The favorable influence of boron on the development of reproductive organs was demonstrated already in 1938 by the investigations of E. V. Babko and V. V. Zimmermann. It was subsequently confirmed by many authors, and extra-root complementary nutrition of plants by microelements has become an agricultural method of increasing the seed yields of many farm plants, such as alfalfa and beet.

Fertility in grass hybrids was increased by applying boron to sowings of poorly fertile wheat-rye hybrids (V. E. Pisarev and M. D. Zhilkina, 1956). E. A. Britikov and R. N. Petropavlovskaya (1954) obtained a considerable increase in the seed set of inbred rye by utilizing enzymes and stimulants of the auxin type. We obtained increased fertility, increased grain size and intensified regeneration of growth in the wheat-Agropyron hybrid, Perennial Wheat M-2, by extra-root complementary feeding with boron and manganese (V. F. Lyubimova and N. N. Seleznev, 1958).

Increased fertility as a result of utilizing microelements and physiologically active substances is explained by an improved metabolism and better development of the reproductive organs.

The following works are of interest in this field: Babko (1950), Smith (1942), Tsung Le Loo and Tsung Chen Hwang (1944), S. I. Emsweller and N. W. Stuart (1948), Richard (1951), Brock (1954), Evans and Denward (1955).

In some intergeneric crosses, the seeds which set have a number of considerable defects, as a result of which they cannot germinate under ordinary conditions. In such cases hybrid plants can be produced if the embryos are grown on an artificial nutrient medium. The investigations of E.V. Ivanovskaya (1946 and 1955), which give a detailed account of the technique, and also work by N. V. Tsitsin and K. A. Petrova (1958) are very interesting in this respect.

In addition to the methods indicated, the environmental conditions in each case can be of considerable importance in overcoming the sterility or increasing the very limited fertility of wide hybrids, in particular during meiosis and the postmeiotic development of sex cells, and also during flowering and fertilization. Very little attention has been devoted to this problem up to now. We therefore think it expedient to go into it in greater detail and deal with the results of some of our investigations in this direction.

During our work with wheat-Agropyron hybrids under the direction of Academician N. V. Tsitsin in the Laboratory of Wide Hybridization, we observed that the sterility of hybrid plants can be considerably modified in many cases. Moreover, within the same plant different shoots often vary strikingly in their fertility. As a rule, a greater quantity of seed sets in late-ripening ears than in ears which develop early.

Cytological investigations of wheat-Agropyron hybrids often reveal marked differences in meiosis in different ears of the same plant.

This gave us grounds to suppose that meiosis in hybrid plants can undergo modifications which depend on the conditions under which meiosis takes place, and which can have a bearing on the fertility of the plants.

The question of modifications in the sterility and fertility of plants is mentioned by numerous authors in the existing literature.

Even Darwin observed that Eschscholzia californica is a self-sterile plant when grown in Brazil, whereas in England it becomes self-fertile. Towards the end of the growing season the seed set suffers a marked decrease. On this basis, Darwin drew the conclusion that temperatures lower than those of Brazil contribute, up to a point, to the self-fertility of this plant, whereas further reductions in temperature lead to a decrease in self-fertility.

Darwin also observed modifications of self-fertility in other plants, such as Abutilon darwini, Thunbergia alata, and Papaver vagum.

Transformations of self-sterile into fertile plants are noted by other authors too. Correns (1916) made such observations in Scrophularia Scopolii, and Baur (1919) in species of Antirrhinum. East and Park (1917) concluded from their work on tobacco that improved external conditions during flowering contribute to an increase in plant fertility. Towards the end of the flowering period, sterile plants become fertile, and, as shown by these authors, the degree of fertility in different tobacco species under identical conditions can be most varied.

Modifications in fertility and sterility depending on weather conditions during different years is noted in fruit plants by many authors: Martin and Iocum (1918), Parrot, Hodgkiss and Hartzell (1919), Wiggans (1919), Welden (1918), Chandler (1918), Gorham (1919), Tafts and Philp (1923), Morris (1921), Auchter (l919-1921, l924), Ryabov (1930). But in all the investigations mentioned, only data of modifications of sterility and fertility are given, whereas the internal changes in the plants which condition their fertility and sterility are not disclosed.

A number of works indicate that high temperatures have an adverse influence on meiosis and so cause sterility of the pollen and egg cells. According to Heilborn (1928, 1935 and 1936), temperatures of above 30°C provoke interruptions of meiosis in some apple varieties and also have a negative influence on the postmeiotic ripening of the pollen, so that it becomes nonviable.

Miedzyrzecki (1933, 1934) reports the adverse effect of high temperatures on meiosis, and the increased pollen abortion due to it in some diploid apples. Nakamura (1932) observed the same phenomenon in Impatiens Balsamina.

Bleir (1930) describes alterations in meiosis, due to high temperatures, in the soft winter wheat Wilhelmina. Pao and Li (1948) deal with modified meiosis as a result of high temperature in wheat, rye, barley, vetch, and other leguminous plants. Elliott (1955) indicates that high temperatures can lead to a decrease in the number of chiasmata in plants and insects. The same was observed by Shams Ul-Islam Khan (1956).

It must be pointed out that modified meiosis and pollen fertility resulting from high temperatures were not connected with a possible hybrid origin of the organism studied in any of the investigations mentioned. This is a most important point, as will be seen from the experimental data given below.

With the aim of finding reasons for alterations in the degree of sterility and fertility in hybrid plants, we investigated meiosis and, later, the pollen dehiscence of the anthers and the grain content of the ears in wheat Agropyron hybrids and their parental forms, which were kept under various conditions.

Investigation of the first generation of wheat-Agropyron hybrids, obtained from crosses of soft and hard wheats with Agropyron elongatum and A. glaucum, showed that the fertility of different ears within the same plant may vary, depending on the conditions under which the sex cells are formed and flowering takes place. If the weather is hot and dry during meiosis, many disturbances are observed, whereas in moderate air temperatures (17 to 20°C) meiosis proceeds normally. For example, in F1 hybrids Tr. vulgare X A. elongatum where somatic cells have 70 chromosomes (21 from wheat and 35 from Agropyron elongatum), 24 to 28 bivalents may be formed, and the configurations during all the stages of meiosis may be normal (Figure 1); as a result, up to 70% of viable pollen grains may develop.

Anthers from the same plants which are fixed in periods of hot and dry weather when daytime temperatures at ear level reach 35° to 37°C may present a very different picture (Figure 2).

Many disturbances occur in meiotic divisions on such days. The number of bivalents varies between 12 to 20, and the remaining chromosomes (26 to 32) are in the form of univalents, causing many irregularities when diverging to the poles. Many of them are fragments and lagging chromosomes. Subsequently many of the univalents fail to become incorporated in the nuclei, and instead form numerous micronuclei. Moreover, chromatid bridges were observed during anaphase. Instead of tetrads, groups of five or six were formed, resulting afterwards in empty abortive pollen.

These observations clearly indicate that temperature changes may induce modifications in meiotic divisions and the subsequent formation of sex cells. Such modifications are decisive for the degree of sterility and fertility of the hybrids.

Investigations of the parent forms of wheat and Agropyron showed no disturbances during meiosis, even at a temperature of 37°C and an air humidity of only 30%. In our opinion this is explained by the fact that, in phylogenetically old forms, the processes connected with reproduction are so well regulated and stable that they take place even under unfavorable conditions. But in the first hybrid generation with its exceptionally unstable sporogenetic processes, there is a very strong response to modifications of the environmental conditions and physiological state of the plants.

Thus, the degree of hybrid sterility depends not only on the phylogenetic proximity between the crossed forms, but on environmental conditions in which meiosis, the formation of sex cells, and flowering take place.

Subsequently we investigated Perennial Wheat M-2 and its F1 hybrids which had been obtained by crossing M-2 with Winter Wheat 2453 and Winter Wheat-Agropyron Hybrid 599.

Perennial Wheat M-2 is the representative of a new 56- chromosome variety of wheat. It is a morphologically and biologically stable variety if guarded against pollination by other wheats. It flowers either openly or cleistogamically, and the grain number per ear varies greatly, even on the same plant; for example, the seed set may be anything between 21 and 73% of the normally developed flowers in an ear.

The first generation hybrids obtained by direct crosses and backcrosses of Perennial Wheat M-2 to varieties of winter wheats Moskovskaya 2453 and Wheat-Agropyron Hybrid 599 are hereditarily uniform within each combination, as was evident in their development, growth, morphology, and other distinguishing. characters. These hybrids are sterile or poorly fertile, like interspecific wheat hybrids.

Meiosis in the F1 hybrid Triticum vulgare x Agropyron elongatum at a temperature of 20°C
Meiosis in the hybrid Triticum vulgare x Agropyron elongatum at a temperature of 35 to 37°C

The experimental plants were grown in large flower pots under identical conditions. At the beginning of June the plants were divided into four groups and placed under different conditions of temperature, air humidity, light and nutrition.

The data recorded in two groups of plants, which were exposed to differing temperatures and air humidities, are most interesting. The first group of plants was placed on an open sunny platform, the second under a humid tent of several layers which was constantly moistened with water.

During a spell of hot, dry weather which set in on the 8th of June, the plants of the first group were subject to the effect of a high temperature of up to 35 to 37°C in daytime, and an air humidity of 30 to 35%. This coincided with the beginning of flowering in the lower ears, and with meiosis in the upper ears.

So as to determine the effect of high temperatures on meiosis, we fixed flowers from the lower ear layer of both groups of plants, and in order to obtain more accurately comparable data, we only used the lower flowers of the eight central spikelets in each ear.

We were thus able to determine differences in the course of meiosis in the different groups of plants. In the first plant group, which was exposed to a high temperature (35 to 37°C), the meiotic divisions in the spore-mother cells showed numerous irregularities. During metaphase there was less pairing of chromosomes than in plants of the second group, which stood in the humid tent at a temperature of 20 to 25°C. In this second group of M-2 plants, the number of bivalents was almost invariably 28, and all the meiosis configurations were normal. At the same time there were mostly 24 to 25 bivalents in plants of the first group, which had been exposed to a high temperature, and in some cases only 21, with the other 14 chromosomes remaining univalents.

In M-2 X Moskovskaya 2453 hybrids, which have a somatic chromosome number of 49 (28 from the perennial wheat and 21 from winter wheat), the number of bivalents varied widely; in the first group there were mostly cells with 14 to 16 bivalents, and in the second there were cells with 21 and sometimes even 22 bivalents.

The number of paired chromosomes had a direct influence on the chromosome distribution in the daughter cells and on the subsequent course of meiosis and formation of sex cells (Table 1).

When meiosis of Perennial Wheat M-2 in the first group of plants took place at a temperature of 35 to 37°C, the anaphase was normal in only 13% of the cells. Anaphase configurations in which there were three to four laggards and fragments, and in which the number of such chromosomes sometimes reached 15 and more, were the largest group (31%). At the same time, the majority of anaphase configurations (67 %) was normal in the second group of plants, where temperatures ranged between 20 to 25°C, and only 7% of the cells had three to four chromosome fragments.

This was found to be the rule in the hybrids. With high temperatures there was a considerably greater number of fragments and laggards, and with temperatures of 20 to 23°C there were fewer abnormal division configurations.


Plants Groups Laggards and fragments
0 1-2 3-4 5-6 7-8 9-10 11-12 13-14 15 and more
M-2 (2n = 56) I (35-37°) 13 22 31 11 5 5 1 3 7
Ditto II (20-35°) 67 21 7 3 1 - 1 - -
F1 M-2 x 2453 (2n = 49) I (35-37°) 0 0 3 18 24 10 11 13 21
Ditto II (20-35°) 0 7 11 17 53 8 3 1 0

There were 21 bivalents in Winter Wheat 2453, as in earlier investigations, i.e. , there was complete pairing of chromosomes of the dividing cells even under the most unfavorable conditions. In the wheat grass Agropyron elongatum 35 bivalents were observed, in A. glaucum 21 bivalents, and all the meiosis configurations were normal.

The differences in meiosis between the two groups of plants (M-2 and its hybrids) also had a direct bearing on the number of normally developed pollen grains in the anthers (Table 2).

In all plants of the second group the number of normal anthers was considerably larger than in plants of the first group growing on the open sunny platform. Here also the negative effect of high temperatures on microspore formation was clearly evident. Among the well-stained pollen grains in the first group there are pollen fragments which stand out by their large size. These pollen fragments apparently derive directly from bivalent chromosomes (dyads) through lack of a second division under the influence of high temperatures.


Plant Medium Maximum Minimum   Medium Maximum Minimum
M-2 22.9 43.0 58.0   58.0 87.3 37.7
F1 (M-2 x 599) 8.0 19.1 30.9   30.9 43.0 17.2
F1 (599 x M-2) 11.3 21.3 33.1   33.1 48.0 22.1
F1 (M-2 x 2453) 5.5 13.7 30.0   30.0 39.3 15.0
F1 (2453 x M-2) 8.1 21.0 32.1   32.1 45.5 13.8

By methodical pollen analyses throughout the entire growing season we established a scheme showing the dependence of the percentage of normal pollen grains in the anthers of the plants studied on air temperatures during meiosis and during the subsequent formation of sex cells.

FIGURE 3. The percentage of normally developed pollen grains in anthers of
Perennial Wheat M-2, and in its hybrids with winter wheats, by air temperature during meiosis.

The curves in Figure 3 represent the percentage of normal pollen in anthers of the studied plants in- relation to the air temperatures during meiosis.

M-2 and its hybrids with winter wheats have similar bell-shaped curves, but in the case of M-2 the peak of the curve is higher and the temperature range is wider. The most favorable temperature for the development of pollen is 18 to 25°C; almost no normal pollen develops at temperatures of about 35°C.

While in the hybrids, too, the maximum amounts of normal pollen are formed at temperatures of 18 to 25°C, these amounts are much lower than in M-2; moreover, at temperatures which slightly exceed 25°C, no normal pollen forms in the hybrids. Low temperatures also have a negative influence on microsporogenesis and cause a series of disturbances in this process. At temperatures of 2 to 5°C, there is a rapid decrease in the amount of normal pollen formed.

Such factors as light intensity, duration and quality, the condition of the assimilation surface, and the presence of nutrient substances in the soil, etc. also exert a considerable influence on gamete formation.

We have not carried out any special experiments to determine the importance of any of these factors, but a few observations may be of interest. For example, perennial wheat plants which were placed in a conservatory in flower pots on well fertilized soil during winter, and which flowered in December to February, contained almost no normal pollen in their anthers; in spite of the high air humidity the anthers remained closed, and no seed formed. In the same plants, shoots which flowered in April and later contained a considerable quantity (50 to 70%) of normal pollen in the anthers. The normal pollen dehisced properly, and as a result of self-pollination the grain content in the ears was relatively high. This clearly illustrates the influence of light intensity and quality and also of the length of the light-day on the process of gamete formation.

The determination of the number of dehiscing anthers showed that there is a direct relationship between the percentage content of normal pollen in anthers and their dehiscence. In Figure 4 the numbers of dehiscing anthers are compared with the amount of normal pollen they contain. In the first group, dehiscing anthers were very rarely encountered. Thus, in M-2 they only amounted to 5.7% of the total, and in hybrids they were found even more rarely or not at all. In the second group, the number of dehiscing anthers was much larger.

A methodical analysis of the pollen of dehiscing and nondehiscing anthers showed that dehiscence varies considerably within the same plant and depends on two causes: on one hand, on the amount of normal, full pollen grains within the anthers, and on the other, on the conditions under which flowering takes place. If it takes place at a low air humidity of 30 to 35%, the number of dehiscing anthers is lower, and in this case even anthers containing up to 60 or 65% of normal pollen may remain closed. However, during moderately warm weather and high air and soil humidity, the number of dehiscing anthers increases markedly, and even anthers which contain only 45 to 50% of normal pollen dehisce. We believe that the turgor state of anther cells and pollen grains contributes to the dehiscence of the anthers in this case.


Plants Group I Group II
M-2 10.1 50.3
F1 (M-2 x 599) 1.0 19.0
F1 (599 x M-2) 2.4 23.0
F1 (M-2 x 2453) 2.9 16.0
F1 (2453 x M-2) 1.6 17.0
Average in hybrids 2.0 18.8

FIGURE 4. Percentage of normal pollen in dehiscing anthers

Differences in meiosis are also evident in the grain content of the ears, i. e. , the number of grains per ear in relation to the total number of fertile florets (Table 3).

In ears of the first group, the temperature during meiosis was 30 to 35° C; the grain content in the ears was very low; in M-2 it amounted to 10.1% and in the hybrids to an average of 2.0%. In ears for which temperatures during meiosis were between 20 and 25°C (group 2) on the other hand, the grain content was much higher. In M-2 it was 50.3% and in the hybrids it averaged 18.8%.

We are justified in drawing a definite conclusion from the results obtained. In many hybrid forms the processes of micro- and macrosporogenesis are subject to considerable variations, depending on the conditions of the environment in which they take place. Changes of environmental conditions lead principally to modifications in pollen sterility of the hybrids. Under the same conditions no modifications in meiosis and fertility are observed in old species.

Until recently, no attention was paid to these phenomena, since they are not always very clearly evident. This is because meiosis in hybrids is apt to be irregular, and one has to analyze large amounts of material to determine the degree of these irregularities. However, the fact that modifications of meiosis under the influence of high temperatures are observed in most cases in fruit plants, and in particular in apples, which are hybrid plants, confirms our conclusions. These are that the processes of meiosis in hybrids, which are relatively less stable than the same processes in any other forms, are easily modified with alterations of the conditions under which they take place.

By providing optimum temperatures and other optimum environmental conditions during meiosis, the postmeiotic development of the sex cells, and during flowering and fertilization, one can in many cases raise a very low fertility and sometimes even overcome the sterility of wide hybrids.

The material which we have discussed briefly shows that there are different means for overcoming the sterility and increasing the fertility of wide hybrids. Further investigations in this direction, based on the achievements of biology and making use of the most up-to-date methods, will undoubtedly enable us to overcome hybrid sterility with even greater success. This will widen the scope for obtaining hybrids between even more remote plants and for creating valuable new plant species.