Wide Hybridization in Plants (1960)
Wide Hybridization in Alfalfa and Other Leguminous Perennial Forage Grasses
E. N. Sinskaya

Doctor of Biological Sciences
All-Union Institute of Plant Cultivation

Most of the alfalfa grown throughout the world is of a hybrid nature. Centers of pure nonhybrid forms of cultivated alfalfa are only preserved in countries with an ancient agriculture in Central and Hither Asia (Afghanistan, Central Asian republics of the U. S. S. R., and Armenia within its ancient limits) and in some oases of the Mediterranean area, though here too the areas under nonhybrid alfalfa are restricted.

In the majority of alfalfa-growing areas the commonly used varieties are products of natural hybridization between only two compound species of medic; blue alfalfa (Medicago sativa sensu lato) and yellow alfalfa (M. falcata sensu lato). Other species are so far only little used in breeding, and varieties obtained by way of experimental hybridization do not generally occupy extensive areas as yet.

Despite the fact that, during recent decades, hybrid alfalfa varieties have been obtained with some success in the U. S. S. R. through interspecific hybridization (works of P. N. Konstantinov, A. M. Konstantinova, I. M. Karashchuk, P. A. Lubents, and others) a sound theoretical basis for production methods of such hybrid varieties is wanting. Methods of obtaining experimental polyploids are being adopted progressively in Western Europe. This means of creating new stocks of starting material for breeding is particularly promising in plants whose vegetative organs have an economic importance - leaves, roots, tubers and bulbs. Varieties of tetraploid Brassica conpestis [campestris?], for example, are already in production in Scandinavia.

There is a need for the development of investigations on obtaining experimental polyploids of alfalfa. The view that in this way positive results can be reached is supported by the existence of very numerous polyploid (tetraploid) medics in nature. Even if experimentally obtained polyploid forms turn out to be of no practical value in themselves, they may greatly facilitate the obtaining of hybrids between certain species, which ordinarily differ in their chromosome numbers and intercross with difficulty or have not been crossed at all until now.

The crosses from which it is most difficult to obtain fertile hybrids are those between species of different chromosome numbers: diploids (2n = 16) and tetraploids (2n = 32).

Among the blue-flowered diploid (2n = 16) medics in the U.S.S.R., the most widespread is the azure medic M. coerulea (a very polymorphic species) and M. hemicycla.

Among the yellow-flowered diploid species, the north-Caucasian M. quasifalcata, the steppe species M. romanica, and the northern M. borealis occupy large and considerable areas. Very interesting is the cold-resistant, high-mountain species M. dzhawakhetica. Not so long ago, the Canadian investigator Lesins (1956) succeeded in crossing M. dzhawakhetica with blue cultivated alfalfa; this opens possibilities for the utilization of this medic species in breeding.

For hybridization, Lesins used a tetraploid form of M. dzhawakhetica, obtained by the action of colchicine on a medic population. The hybrid obtained was almost self-sterile, but set seed when backcrossed to tetraploid blue cultivated alfalfa.

The obtaining of forms with artificially induced tetraploidy from diploids is one of the means by which fertile interspecific hybrids between medic species with different chromosome numbers can be produced; undoubtedly, the use of such forms will aid the hybridization between other diploid and tetraploid species of Medicago.

Many foreign investigators show a considerable interest in our diploid wild-growing species. Oldemeyer and Brink (1953) have established that diploid M. falcata hybrids (from Orenburg) preserve certain hereditary peculiarities and are more winter-hardy than the known northern alfalfa varieties. Karlis Lesins (1954) found that "diploid yellow alfalfa from the northeastern Caucasus" (it was probably M. quasifalcata) is very resistant to bacterial disease (wilt); this is confirmed by our observations (1948).

Scientists in Canada, the U. S. A., Scandinavia, and other countries are very interested in the U. S. S. R. diploid yellow medic as an object for theoretical investigations and practical breeding, though they have only very scanty and randomly obtained material at their disposal. Unfortunately, the inexhaustible resources of immensely variable diploid yellow medics in the U. S. S. R. are almost ignored by the breeders and have so far not been sufficiently studied geobotanically.

A number of foreign scientists also study the occasional diploid forms which are rarely encountered in stands of cultivated blue alfalfa (for instance in Grimm alfalfa in America). Such forms are conveniently used for obtaining fertile hybrids with diploid wild-growing blue- and yellow-flowered medic species.

Already in the pre-war years we succeeded in obtaining some fertile hybrids between medic species of different chromosome levels through pollination with a pollen mixture (1941, 1950). In recent years, this method has been tested and worked out in more detail by V. A. Borkovskaya (under our direction) at the Maikop Experimental Station of the All-Union Institute of Plant Cultivation. The results of her investigation can be summarized as follows. In crosses between diploid and tetraploid medic varieties entirely fertile hybrids can be obtained by pollination with a suitably prepared mixture of pollen from a number of different donors, instead of the sterile or poorly fertile triploids which were formerly obtained as a result of pollination with pollen taken from only one particular seed parent. In a number of cases, yellow diploid medic was most effectively pollinated with pollen from plants of various tetraploid ecotypes (considered separate minor species by some taxonomists) of cultivated alfalfa (for instance the Hither Asian, Armenian, Tripolitanian, and other ecotypes). If a diploid is taken as seed parent, more pods are formed, but there is no difference in the quantity of obtained seed between hybrids from diploid or tetraploid seed parents.

In crosses between cultivated and wild-growing forms, it is advisable to take the latter as seed parent; the ensuing hybrids may then exhibit a higher resistance to diseases, damage and unfavorable conditions. For practical breeding purposes the cross can be made without previous emasculation of the maternal flowers - the percentage of seed set in the pods is thus higher than with emasculation.

The seed yield in diploid M. coerulea pollinated with a mixture of pollen collected from different forms of cultivated tetraploid alfalfa was four to seven times that in plants for which the pollen donor was only one single alfalfa. By using this method, fertile hybrids are easily obtained from species which are not easily crossed in the usual way. For instance not a single hybrid was obtained by investigators who pollinated thousands of emasculated flowers of Medicago quasifalcata with pollen taken from one particular donor, whereas the use of a pollen mixture led to the obtaining of hybrids without any great difficulty.

Natural hybrids between the diploid and tetraploid medic species which we have studied are all fertile allotetraploids; undoubtedly, their production is due to insect pollination of the racemes by a pollen mixture. Hybridization between triploid and tetraploid species is very often more successful (in natural as well as artificial pollination) in districts which are at a distance from the main area of seed-parent distribution. I. M. Karashchuk, for instance, observed that crosses involving M. quasifalcata are accomplished more easily in western Siberia (Barnaul) than near Maikop.

This may be explained by the fact that incompatibility in nature is a means of preserving the entireness of the species. Our observations have disclosed a number of instances where the boundaries of distribution areas of diploid and tetraploid species are contiguous with each other, but where incompatibility between the species is maintained by a physiological barrier (different chromosomes levels), so that no intermediate hybrid forms are found in these border-line areas. In the forest-steppe and partly in the steppe zone of the northern Caucasus, for example, the common tetraploid yellow alfalfa M. falcata sensu stricto does not intercross with the diploid M. romanica, in spite of the adjoining boundaries of the two species' distribution areas. In the same way, M. quasifalcata is isolated by a physiological barrier in the west of its distribution area from tetraploid M. glandulosa, and in the south from tetraploid M. falcata sensu stricto.

This incompatibility is determined by a selection under the particular conditions on a given boundary area, and the percentage of incompatibility is quite often observed to vary with varying conditions in different districts.

If species which natural migration has brought from afar enter conditions which are very different from those of their earlier areas of distribution, the tendency of two species to mutual incompatibility may weaken, and then one can observe a mass production of natural interspecific hybrids between species of different chromosome levels in certain localities; under such conditions, not triploids, but fertile tetraploids are formed (we have studied an occurrence of this kind in Daghestan).

Thus, this method of producing fertile allotetraploids through pollination on with a pollen mixture is a natural and very efficient means of obtaining fertile hybrids between diploid and tetraploid species; it should not only be used in connection with alfalfa, but also in other plants.

However, this question has not yet been treated sufficiently from the genetic and the cytological points of view. The processes of fertilization and embryo development in the formation of fertile allotetraploids have not been studied. There are no grounds for the assumption that this is a case of pollination by unreduced pollen of the diploid pollen parent, as most of our best hybrids were obtained by pollination with pollen of tetraploid cultivated forms.

Ledingham (1940) expressed the opinion that, in such cases, unreduced egg cells play a part in the formation of fertile hybrids between diploids and tetraploids; however, no factual evidence is available on this question.

The application of Michurin's methods should by no means exclude the clarification of these questions from the genetic and cytological points of view. Interspecific medic hybrids, in particular those obtained through crosses between species of different chromosome levels, are of considerable theoretical and practical interest.

In the production of heterosis forms, stress should be laid on the quantity of herbage and grain productivity. It has been shown in Denmark that particularly marked heterosis was exhibited by beetroot hybrids between diploid and tetraploid forms. The same phenomenon was observed by us in interspecific medic hybrids.

Heterosis in vegetative mass in alfalfa is usually related to a more complex structure of the plant, increased "bushiness" (number of main lateral shoots) and branching (number of "side branches", i. e., sublaterals and third laterals), and a higher leafiness (which is connected with a decrease in the length of internodes and an increase of their numbers along the shoot).

According to the investigations of I. M. Karashchuk (1955) and ourselves, heterosis in vegetative mass in hybrid alfalfa populations accumulates gradually in the course of several generations as the architecture of the "bush" becomes more complex. Such is the situation in hybrid populations obtained from natural crosses between medic species.

Extensive investigations of interecotype hybrids of cultivated alfalfa, carried out under our direction at the Maikop Station of the All-Union Institute of Plant Cultivation, have shown that vegetative-mass heterosis is exhibited in many combinations already in the first generation. Mediterranean ecotypes (Tripolitanian, Yemenite and Greek alfalfa) proved to be particularly valuable, as they have a tendency to exhibit marked heterosis in combination with various components. Unfortunately, such ecotypes, which manifested a particularly marked hybrid vigor, have not been preserved in the Institute's collection. Steps should be taken to restore this valuable stock of alfalfa forms exhibiting heterosis.

Interecotype hybridization in alfalfa also belongs to the category of wide hybridization, as the best results are usually obtained upon crossing geographically remote forms (part of which are described by taxonomists as cross-pollinated in pairs or groups, very variable results are obtained. The combining ability of each particular biotype must be established for the different conditions of the agricultural environment. An even greater variability of biotypes will almost certainly be discovered in complex hybrid populations of medic, which constitute a pool for the guided selection and cross-pollination between groups; this need not lead to any harmful impoverishment in the variety composition (this has been demonstrated, if only in the case of sunflower, in the breeding successes of V. S. Pustovoit and others).

Willful selection contributes to the development in plants of those quality characters which need not be essential biological attributes and are not necessarily closely linked with the viability of the organism. Examples of such characters are nonshattering, difficult pod dehiscence in alfalfa, the firm attachment of pods in sainfoin (Onobrychis), the specific chemical composition, etc.

* [Kul'turnaya flora SSSR.]

Breeding for fodder quality and a definite chemical composition is a very urgent problem. A paramount need in alfalfa and other fodder herbs is the increase of protein and vitamin content. To satisfy these needs interspecific and inter-ecotype hybridization should be resorted to. According to investigations by Smirnova-Ikonnikova (see "Flora of Cultivated Plants of the U. S. S. R.*" Vol. XIII) higher quantities of proteins and vitamins can accumulate in certain hybrids than in the respective starting forms grown under the same conditions.

The chemical composition of the starting forms can be determined from average samples only where one deals with large-scale open cross-pollination of whole interspecific or inter-ecotype hybrid populations. For more thorough work every parent plant is used directly and its combining ability in group cross-pollination is tested afterwards.

To study the chemical composition of parent plants, the latter have to be propagated vegetatively. Alfalfa is propagated very easily from parts of the "bush" and cuttings. If cuttings are used, they have to be taken as far as possible from the same layer of laterals or from closely adjoining ones. If cuttings from the lowermost branching level are taken together with others from the top level, the clones may lack biological and chemical uniformity. It is most advisable to take cuttings from the lower branching levels.

Cuttings can be taken repeatedly, until the number of clones is sufficient for a precise assessment of the starting plants' chemical composition and other properties; there should be enough material for an adequate number of replications and for analyses of chemical composition under differing conditions. After cross-pollinating the clones in pairs and in groups, their combining ability is determined.

Our breeders have somewhat forgotten the ease with which medics can be propagated vegetatively, and in the U. S. S. R. this valuable property is little exploited at present. In a number of foreign countries, on the other hand, clones are widely used in breeding. In Denmark, for instance, the selected plants are propagated vegetatively and are sown or planted in groups for cross-pollination (100 plants of every clone). Cross-pollination in the desired direction is effected by separate pollen collection from each clone, and by cross-pollinating in pairs or in groups with a mixture of the pollen; thus large numbers of homogeneous hybrids can be obtained, which can themselves be cross-pollinated as required.

The method of Levin and Pedersen (1955), who obtained pollen from separate clones, should be tested. For this, bees and bumble bees were used, which had to collect pollen from one clone (or from a definite group of clones). The pollen was collected from the bees by special catching appliances fixed to the entrances of the hives. Through directed cross-pollination between groups with the utilization of clones, uniform complex hybrid populations can be obtained.

Below we present a scheme for breeding medics which uses new methods based on the data of Canadian and American breeders. This scheme can be utilized for selections within an ecotype or variety population, and in crosses between two ecotypes or two species (of different chromosome levels).

In the latter case, the clones are planted in the populations of both species, and interspecific hybrids are obtained from pairs of analogous clones (from parallel ecotypes and eco-elements) as well as from non-analogous ones. The latter hybrids sometimes exhibit a more marked heterosis, while adaptability to a definite complex of conditions of the agricultural environment can be increased considerably in hybrids of the first category (where the starting forms are ecologically analogous).

Scheme for alfalfa breeding using pairs of hybrid clones and the study of combining ability of the clones in polycrosses and top crosses.

  1. Choice of stock planes for clonal division from the starting population.
  2. Planting of clones (vegetative propagation of the selected plants); determination of the degree of self-fertility of the clones; rejection of plants with the highest self-fertility.
  3. Crossing of pairs of clones by planting alternate rows on isolated plots.
  4. Study and rejection (selection) of pairs of hybrid clones; determination of dominance, heterosis and the influence of maternal heredity-assessment of each clone separately; determination of specific combining abilities.
  5. Testing the selected clones in top crosses with variety B, which is specially chosen for this purpose (white-flowered, if all clones have colored flowers); the clones and variety B are planted on an isolated plot in alternate rows.
        An additional examination can be the testing of all chosen clones by planting them in alternate rows with variety B on an isolated plot.
  6. Testing the clones in polycrosses by open cross-pollination between each clone and a large selection of the best clones (previously studied as well as new ones) on isolated, differently exposed plots (determination of the general combining ability).
  7. Development of synthetic populations by open cross-pollination between the best selected clones on isolated plots.
  8. Comparison between the best synthetic populations for obtaining a synthetic variety.

All we have said also applies in some measure to other leguminous herbaceous perennials, particularly to sainfoin. As compared with alfalfa breeding, the enormous variety of sainfoin species is as yet little used in breeding work. Combining abilities are being tested, but this has been done in only a few of the main species. The majority of the most promising species of subsection vulgatae intercross well and often form natural hybrids. But here surprises are encountered, which dictate the necessity of an experimental verification of compatibility between the species. Among the north-Caucasian mountain species of sainfoin, for instance, which have been studied by us and by V. A. Borkovskaya in the neighborhood of Gelendzhik, Onobrychis miniata did not form hybrids with O. Grossheima; this would have been a desirable cross in which the resistance to soil drought and rust of the first species might combine with the growth vigor of the second.

At the present, only Onobrychis arenaria sensu lato, O. viciaefolia, and the Transcaucasian sainfoin are being used in interspecific hybridization. The latter is not a real botanical species, and represents a complex population of plants belonging to two species - O. a1tissima and O. transcaucasica, and hybrids between them. In the Akhalkalak District of Georgia, a third species, O. Biebersteinii enters into the composition of this population.

This last high-mountain species is valuable for its cold resistance. Moreover, it is marked by polymorphism, and very valuable ecotypes can be selected from it, which are frequently promoted by taxonomists to the rank of species. They have a good leafiness and yield considerable amounts of cutting mass. O. Biebersteinii is used by N. P. Dmitriev in breeding by the hybrid method at the Stavropol' Territory Animal Breeding Station.

Even less has been done on the interspecific hybridization of clover. Until recently, the view prevailed that species of perennial clovers do not intercross, and that therefore most of them cannot be used in breeding. However, the work of the Ural breeder D. M. Savchenyuk has shown that interspecific hybridization of clover is entirely possible. He obtained hybrids between red Trifolium pratense and pink T. hybridum. In this way, D. M. Savchenyuk obtained new clover varieties which are more resistant to lodging and richer in nectar.

A further thorough genetic ecologic study of clover species will undoubtedly lead to new breeding successes by interspecific hybridization.