Wide Hybridization in Plants p 301-307. (1960)
WIDE HYBRIDIZATION IN THE MALVACEAE

K. A. Vysotskii
Central Asian Experimental Station of the All-Union Institute of Plant Cultivation

The development of high-quality, better yielding, new cotton varieties which ripen early, are disease resistant, adapted to local conditions, and suitable for mechanized cultivation and harvesting, is an urgent item among the measures to be taken to increase bulk production of raw cotton. There is also, as yet, a shortage in some of the important cotton districts of the Soviet Union of quick-ripening varieties, which have no exacting temperature requirements during the germination period and are resistant to the Epitetranychus mite and to fungus diseases.

In former large-scale hybridization projects carried out at the Central Breeding Station of the All-Union Cotton Research Institute, we made successful crosses involving the four cultivated cotton species: Gossypium hirsutum L.; G. barbadense L.; G. arboreum L., and G. herbaceum L.) and also representatives of three wild species (G. Thurberi  Tod., G. anomalum Wawra et Peyr. and G. Sturtii F. Mull.). In the interspecific hybrids obtained, in particular in those from the cross between G. hirsutum and G. barbadense (Nos. 148 and 138), economically valuable characters appeared in quite a favorable combination — lint length (37 to 38 mm), boll size (up to 7 g) and good fruit bearing; yet these hybrids were not taken up in production, because of their insufficiently quick maturation.

Contrary to expectations and despite work in this direction throughout a whole decade, no positive results were obtained at that time from quite a large group of haploid, heteroploid and polyploid cotton forms, in which the chromosome numbers were 2n = 26, 39, 65 or 78.

In 1934, we obtained the first sexual hybrid from a cross between upland cotton G. hirsutum L. and the wild Alcea rosea.  The plant was marked by intermediate morphological characters; it flowered abundantly, but proved entirely sterile.

*Apparently, Malva
and Althaea species

The idea of the creation of hybrids by crossing cotton with other genera of the mallow family is very appealing. The genera Hibiscus and Alcea comprise forms which are valuable with regard to productivity, early ripening, resistance to low temperatures and heightened soil salinity, and immunity to diseases and pests. In mallows*, fruits mature within 22 days from flowering, and in Hibiscus within 22 to 28 days, whereas the flowering-to-maturation period in cotton plants amounts to 50 or 60 days and more.

Cotton, Hibiscus and mallows also differ considerably from each other in the length of the first vegetative stage the period between germination and bud appearance. In Hibiscus this period is 14 to 15 days shorter than in cotton. In all the commonly grown industrial cotton varieties, 10 to 16 sympodial branches are present on each plant at the beginning of ripening, or as many as 20 where growing methods are on a high level. At the end of the growing season, after the natural ovary drop, ten to thirty bolls remain on the plant and provide for a raw-cotton harvest of 25 to 50 centners/ha.

In mallows, as against cotton, there is a considerably lower percentage of ovary drop, sometimes no ovaries dropping at all; moreover, the racemose fruiting inflorescence is constituted of a much larger number of fruiting branches. If growing measures are on a high level, up to a hundred fruiting branches may form in a plant and 3,000 to 5,000 fruit elements; this is many times the number of fruits formed in cotton.

The high polymorphism and the enormous geographical distribution area of Hibiscus and the mallows point to the exceptional adaptability of species of this family to the conditions of different habitats. These peculiarities of the Malvaceae furnish the breeder with great possibilities in the choice both of starting material for the cross and of the subsequent varieties obtained, in order to get forms which are best adapted to local conditions.

When starting work in this direction, we encountered an apparently serious obstacle: the high incompatibility between different genera and tribes of the Malvaceae, in particular, between the different genera and cotton. To overcome this obstacle, we resorted to a number of new methods; firstly, multiple saturating pollination. A pollen mixture from different genera of the Malvaceae was applied three times and, subsequently, pollen from the maternal plant was used. Sometimes these operations were effected on the same pair of parental forms over three, four and even seven years. All the buds on the cotton seed parent, with the exception of the emasculated ones, were removed, creating more favorable conditions for the survival of the pollinated ovary. Pieces of the pistil of the pollen parent were placed on the stigma of the emasculated flower of the seed parent together with the pollen. We also used other methods of acting on the generative tissue.

The most diverse mallow and Hibiscus species which were involved in hybridization helped to make the crosses successful. The modified crossing method, in which the pollen of Malva and Hibiscus was used as sexual mentor to modify the metabolism, aided in the solution of two basic problems: first, to obtain entirely new hybrid forms with characters new for the genus Gossypium, secondly, to induce these forms to bear fruit.

Particular attention should be paid to the fertile forms which we obtained from the cross between upland cotton G. hirsutum  and representatives of the weed vegetation (Alcea rosea, Malva neglecta, Malva  pulchella), ornamental plants (Hibiscus Moscheutos, H. coccineus, the swamp mallow of North America) and fiber plants (H. cannabinus, H. esculentus).

The first-generation hybrids proved to be most original. The F1 G. hirsutum (2n=52) x Hibiscus coccineus Walt. (2n=38) is a dwarf, many-branched, bushy plant, which develops up to eight stems. The height of each stem does not exceed 45 cm. The leaves of the hybrid have three to five elongate, narrow lobes arranged on complementary, independent petuoles, give the impression of pinnately divided leaves absolutely untypical of the mallow family. The hybrids' general branching habit and shape has been inherited from the pollen parent H. coccineus. According to the preliminary data of the cytological analysis, 45 chromosomes are present in the somatic cells of the hybrids: this is in accordance with expectations.

First-generation plants of G. hirsutum x H. esculentus (okra) differ markedly from the former hybrid by the considerable size of the whole plant and the abundant leafiness. Their height is 120 cm. The large leaves are weakly cleft and the leaf blade surfaces are bloated and have semicrenate margins. All the morphological characters of the hybrids, with the exception of the height of the main stem, are obviously derived from the pollen parent H. esculentus; this also applies to the external structure of the very markedly ribbed, elongate capsule. Most of the plants are sterile, and only rare specimens are weakly fertile, with a very late period of development.

The second type of hybrid plant of the same combination, the F1 of G. hirsutum x H. esculentus, differs from the first by an even less divided, very wide leaf blade with inturned lobes. On the whole, the bush has an abundant leaf cover and a firm upright stem. The leaves are very coarse, hard, and have almost crenate margins. All these characters have been inherited from the pollen parent.

The F1 of hybrids G. hirsutum x Alcea rosea, with semicircular or circular leaves and a specific mallow-type growth habitat, are rather original, with a growth habit of an intermediate character. The flowers are small. A number of capsules with normally developed seeds (up to five) developed in some hybrid plants, particularly in those of the second generation, upon pollination with pure pollen of the paternal form (A. rosea), whereas plants pollinated with cotton pollen dropped their ovaries at an early age.

The first and second hybrid generations of G. hirsutum x H. cannabinus closely resemble the seed parent H. cannabinus in their general growth habit, leaf structure and arrangement of flowers and ovaries (Figure 1). The distinguishing trait of these hybrids is the structure of the ovaries and capsules. There are four to six involucral bracts in the hybrids, instead of the three characteristic for the entire genus Gossypium. This character in the hybrids is also derived from the seed parent (Hibiscus cannabinus). In single hybrids obtained by pollination with pure H. cannabinus pollen, a number of small capsules with normal seed were formed.

In addition to the specific segregation of morphological and quantitative characters depending on the heterozygosity of the parental forms of Malva, Alcea and Hibiscus, first generation cotton-mallow and cotton-Hibiscus hybrids also vary in their fertility. Data collected over a number of years indicate that the earliest ripening and most productive hybrids are those from the crosses G. hirsutum x Alcea rosea, and that the most fertile hybrids derive from the combination G. hirsutum x Malva neglecta, and others. Among the latter, sterile forms represent a most insignificant percentage. As a rule, capsules set in these hybrids even if only in small numbers from fertilization by pollen from the paternal form or from other fertile hybrids. At the same time, there are other forms with a very high fertility, in which some hundreds of capsules per plant develop. In certain F2 and F3 hybrid plants, capsules set upon pollination with pollen of both cotton and mallow.

FIGURE 1. F2 Gossypium hirsutum x Hibiscus cannabinus.

The F1 hybrids between G. hirsutum and representatives of the genus Hibiscus have different properties, particularly the combination G. hirsutum x H. cannabinus. The fertility of this hybrid group is notably lower than that of the preceeding one. There are more sterile or very weakly fertile plants in this group (70 to 75%). There are few plants in which self-fertilization occurs because of a high percentage of nonviable pollen and reduced egg cells. The artificial pollination of hybrids of this group with Gossypium pollen does not result in more than 7 to 10% of ovaries. Upon open flowering, capsules form very rarely in these hybrids, and then only at the age of two years.

Some constant lines have been disclosed among cotton-mallow hybrids in the third and often also in the second generation. Such lines recombine certain morphological and physiological characters of both parent forms and also exhibit new characters, unusual for the genus. This demonstrates the intensity of form-origination in interspecific hybridization.

The study of a considerable number of not hitherto investigated F3 to F5 hybrids from various combinations of parent forms revealed a number of important facts. In the first place, there were many new types, exhibiting segregating characters shown by remote wild forms of the genus Gossypium.

* Species unspecified.

One of the remarkable phenomena observed in intermediate hybrids of the first to fifth generations was a considerable regrouping of morphological and physiological characters of the parents. In one instance, in which the starting cotton form No. 1306 (Skorospelka Quick-Ripening) had a relatively low potential yield capacity and did not ripen particularly early, cotton-type plants of the third-hybrid generation (No. 1306 x vegetable-garden mallow*) included forms which were more quick-ripening (they ripened 25 days earlier than No. 1306) and were abundantly fertile. Moreover, capsule formation did not take 50 to 54 days as is characteristic for variety No. 1306, but was fully accomplished within 30 to 35 days from flowering.

Even more valuable results were obtained in the combination No. 0100 (Navrotskii) backcrossed to mallow. Here the capsule reached a weight of 11 and even 12 g, and this high weight was associated with a fiber length of 36 to 38 mm and other high technological qualities (metrical number 6500), as well as with a compact growth habit. A four-year study of this material confirmed its high breeding value. In general, forms inclining to the cotton side are the most productive and, therefore, promising.

We succeeded in selecting a number of families, which transgressed far beyond the limits of the genus Gossypium and which did not segregate in subsequent generations, from the second generation of intergeneric hybrids marked by a complex of new characters. The hybrid nature of these lines was confirmed by an analysis of pollen-grain size. Physiological characters and also the morphology of the hybrid plants indicate a combination of different traits inherited from cotton, Hibiscus and mallow.

In backcrosses of these forms to Hibiscus or mallow, there is rarely a sufficient percentage of successful fertilization, and a subsequent complementary pollination with the hybrids' own pollen has to be resorted to.

In backcrosses to cotton, the percentage of fertilization is sometimes higher.

The progeny of plants exhibiting morphological characters of the pollen parent (i.e., in which characters of Hibiscus or mallow predominate) are very often marked by partial or complete sterility. But entirely fertile forms are also met with among this group of plants. Finally, there are a number of forms which are morphologically distinct from both cotton and other malvaceous species.

Almost all hybrids between cotton (as seed parent) and other malvaceous species exhibit pollen-parent characters and are, as a rule, distinguished by finer pollen grains; they also approach the pollen-parent in citric-acid content (analyses by Professor P. G. Chesnokov, Leningrad).

It must be stressed that it is not easy to combine the most valuable characters of cotton with early ripening and a high resistance to diseases and unfavorable environmental conditions, in other words, with high adaptability. We needed more than ten years of obstinate endeavours before we obtained the first forms approaching a combination of high yield capacity, compact growth habit, early ripening, very large capsules with high-quality fibers, and disease resistance. And we are not yet anywhere near the complete solution of this problem.

In view of the quick ripening of the lines obtained (Nos. XX, 312, 1973, and others) and their possible usefulness in solving the problem of introducing cotton to more northern districts, we began experiments in 1956, in which these forms were used as starting material for breeding varieties with a shortened vegetative period.

To this end, the early ripening Skorospelka (line XX) was sown in open ground on 8 June, i.e., more than one month and a half after the usual date of cotton sowing. The first flowers on these early ripeners sown at this date were formed 45 to 47 days after sowing, i.e., on 23 to 25 July, and the capsules began to open on the 90th day. By the beginning of October, the plants had almost completely caught up in development and had reached a normal state in growth, fruiting branches (11 to 12 sympodial branches per plant), and number of bolls.

FIGURE 2. From left to right: Alcea nudiflora; G. hirsutum x Alcea nudiflora; line 20/2307 an ultra-quick-ripening, resistant cotton variety

In 1957, samples taken from bulk selections of line XX were tested at Isfar (Tadzhik SSR); prior to the frosts, they showed themselves superior to all existing standards in yields.

The obtained forms have 11- and 12-gram bolls and a staple length of 36 to 38 mm. The fairly widely grown industrial variety 108-f has a boll weighing 7 to 7.5 g and a staple length of 31 to 32 mm. Family No. 1973, the best of the hybrid families, has a boll weighing up to 12.9 g and a 37 mm staple length.

The new variety Golodnostepskii 912, which exhibits a complex of positive characters, is very promising: the staple length is 36 to 37 mm, the lint yield is high and the fiber strong, the plant is compact and resistant to lodging. Forms which are resistant to damage by the common pests of the cotton belt, in particular to damage by the Epitetranychus mite, deserve special attention. These hybrid samples have a dark bordeau coloration of all their parts (with the exception of the raw fibers) preserves [sic] the variety from biological and mechanical impurities; the stem resists lodging, the fiber is of high quality, and the seed is naked. This last is of great importance in mechanized sowing and helps to accelerate seed germination.

The new line 2307, obtained from hybrids of G. hirsutum x Alcea nudiflora (Figure 2) is even more interesting in its resistance to Epitetranychus damage, and, at the same time, to bacterial blight (caused by Xanthomonas malvacearum) and wilt. This line is marked by a compact branching habit and a very strong leaf pubescence which, apparently, constitutes an obstacle to the sucking pests of cotton.

Attempts to cross cotton with other representatives of the Malvaceae have been made in many countries and on many occasions, both in the past and the present century. Without exception, all these attempts have failed.

Through the newly adopted method of multiple saturating pollination over many years, using Hibiscus and Malva pollen mainly as sexual mentors modifying the metabolism, profound changes were wrought in the process of form-origination. Hereditary modifications resulted, and valuable character combinations could be fixed within one form.

Experiments have confirmed the existence of an enormous variety of character complexes for early ripening, high productivity, and viability, which are absent in the genus Gossypium. These data testify to a certain phylogenetic proximity between the group of species of the four large genera Gossypium, Hibiscus, Malva, and Alcea, though it is difficult as yet to visualize a clear scheme of the phylogenetic relationship between Gossypium and other genera of the large mallow family.

Further intensification and expansion of genetic and breeding research, with the introduction of new species and genera of the Malvaceae into the crosses, will doubtless produce new results with a theoretical and practical interest.