G.M. Darrow, The Strawberry: History, Breeding and Physiology
THE EARLY HISTORY of the strawberry in Europe and the New World, including an account of its development in England and France up to 1900, appears in the preceding section, along with a review of the literature accompanying the strawberry's development. The early development of strawberry varieties in America and the men who bred them will be described in other chapters, as will later breeding work. First, however, it seems best to discuss Fragaria more fully, in order that breeding work of the past, present, and future can be properly appreciated, both for its accomplishments and the material limitations it must accept. To effect the basis for this appreciation it is necessary to describe strawberry species, both as they are distinguished by their genetic composition and by their gross morphological characteristics, for such description helps to clarify both the sources of characters and the possibilities displayed by the modern strawberry. Once this is done, the later discussion will rest on a broader base of information and will be more useful.
For a clear understanding of the possibilities and limitations of strawberry breeding, the results of studies on strawberry genetics and cytology must be presented. The material which follows contains the findings of various investigators in this field; of Millardet, who obtained maternal and paternal inheritance in certain crosses, the descendant seedlings strongly resembling either the mother or father; of Solms-Laubach, who confirmed Millardet's findings; of Strasburger, who found normal union of egg and pollen in such crosses, but with dominance of paternal characters; of Richardson, who determined that the diploid vesca displays what we term simple Mendelian inheritance, while the octoploid garden strawberry displays a complicated inheritance; of Longley, who first obtained the chromosome numbers of several strawberry species. In such fashion, through the work of single investigators, basic information for breeders was gradually built up. Later investigations of the genetics and cytology of the strawberry were carried out on a considerably larger scale, especially by three centers of research: The Friedrich-Wilhelms University, beginning with Miss Elizabeth Schiemann around 1919 and continuing through her students, especially Dr. G. Staudt, until now; the Bussey Institute of Harvard University from 1921 to 1941, with Dr. East and his students, the Drs. Ichijima, Manglesdorf, and Yarnell; and the University of Manchester from 1950 to 1959, with Dr. Harland and his students, Margaret Smith and Drs. King, T. Smith, Ellis, Islam, and Jones. The studies at these research centers did much to establish the relationships among strawberry species. These studies also indicated the probable ancestors of the modern strawberry, and whether useful new types might be produced with different chromosome numbers, with different compositions of chromosomes from other species, or as hybrids with related plants such as Potentilla. Other genetic research is described under the country and institution where it was pursued, especially under North Carolina and the U. S. Department of Agriculture at Cheyenne, Wyoming, for methods of breeding; the U. S. Department of Agriculture, for production of productive vesca-flavored decaploids; and Russia, for cytogenetic studies.
A. Millardet, Professor of Botany at Bordeaux, France, was really the first modern strawberry geneticist. He began a study of hybrids of Fragaria in 1883 and continued this study eleven years. He published his paper Notes on Hybrids Without Union or False Hybrids in 1894. From the cross vesca alba x chiloensis he obtained one sterile plant similar to vesca and three plants like chiloensis, the pollen parent. In all, he obtained four cases of paternal plants where they resembled the chiloensis in all general characters.
In Germany, H. Solms-Laubach repeated Millardet's work and in 1907 reported that he also obtained only paternal plants from virginiana x moschata, all of which were sterile, as male plants of moschata characteristically are.
Strasburger (1909) sectioned flowers of the virginiana x moschata at twelve-hour intervals after pollination to determine if actual fertilization of the egg by the pollen took place. In the flowers he studied, he found union of egg and pollen typical of sexual reproduction, indicating that the so-called "paternals" were true hybrids, with the characters of the pollen parent remarkably dominant in the seedlings where moschata was a parent.
C. W. Richardson (1914, 1918, 1921, 1923) made a wise choice in the kind of strawberries he selected for his research. In studies of inheritance, he used the everbearing wood strawberry, F. vesca semperflorens, a diploid in which inheritance is clearest and simplest, involving runnered and runnerless, 3-leaflet and 1-leaflet, white and red-fruited forms. He found that the first generation of the cross "runnered x runnerless" plants always produced runnered plants, but when selfed these hybrid plants produced both runnered and runnerless seedlings in good Mendelian fashion. Likewise, 3-leaflet x 1-leaflet gave all normal 3-leaflet seedlings in the F1, but 177 of 3-leaflet to 73 1-leaflet in the F2; white-flowered x pink-flowered gave all pink-flowered in the F1, but 77 pink to 10 white in the F2; double-flowered x single-flowered gave 97 single to 37 double in the F1; and white-fruited x red-fruited gave all red-fruited in the F1 and 70 red to 20 white in the F2. Thus, inheritance in the diploid strawberry was shown to be like that in other plants.
In contrast, when two large-fruited everbearers (octoploids) were selfed, more than a 3 to 1 ratio of everbearers (177 to 33) was obtained, while noneverbearing x everbearing gave 24 everbearing to 53 non-everbearing. All non-everbearing seedlings of this cross when selfed gave at least some everbearers. In crosses of female x perfect-flowered, Richardson obtained 183 female to 155 male, or perfect-flowered, about a 9 to 7 ratio. He obtained one cross with all large fruit, some reaching 32 cubic centimeters in volume. Royal Sovereign selfed gave 12 seedlings with flavor equal to their parent, three with somewhat better flavor, and 96 with no flavor or with a bad flavor. At the end of his research, he decided that it was not possible to judge the flavor of selfed seedlings because their fruit was too small.
When strawberry breeding began with the U. S. Department of Agriculture, at Glenn Dale, Md., the need for information about the chromosome numbers of the species and varieties was soon evident. In answer to this need, A. E. Longley (Fig. 7-1) began his studies about 1923. In Chromosomes and Their Significance in Strawberry Classification (1926) Longley first reported the chromosome number of vesca, moschata, chiloensis, virginiana, and of many varieties and hybrids. He considered vesca to represent the primitive form from which the others evolved, and held that dioeciousness (male and female flowers on separate plants) in the strawberry was associated with polyploidy. He also confirmed Millardet's results in obtaining maternal and paternal forms in crosses. Fig. 7-3c, left, shows the 28 chromosomes of pollen mother cells of the variety V. Hericart de Thury and the seven chromosomes of pollen mother cells on the right of two diploid seedling varieties of vesca.
The report in 1894 by Millardet that some species-crosses of strawberries gave either maternal or paternal seedlings led E. M. East (Fig. 7-2) and his students at the Bussey Institute, Harvard University, to undertake a study of the genetics of Fragaria. Genetic investigations began in 1921 and continued through 1941. In 1930 East published The Origin of the Plants of Maternal Type which discussed maternal-type plants that occurred in interspecific hybridizations. East also described a diploid plant resulting from the cross F. vesca (2n = 14) x F. virginiana (2n = 56). The maternal parent F. vesca had white fruit, while the seedling resembled vesca but had red fruit. East then grew selfed seedlings of the red-fruited F1, and reported the results in 1934 in the paper A Novel Type of Hybridity in Fragaria. Of the 18 greatly variable selfed seedlings raised, three did not flower, seven had vesca-like fruits, one of the seven having seeds in sets like virginiana, and the eight others were entirely sterile. Sixteen were diploid (2n = 14) and two were triploid (2n = 21). East concluded that one set of chromosomes of F. virginiana is so like those of vesca that pairing of chromosomes occurs and plant development follows.
In 1924 K. Ichijima began a study of the chromosome number of the species and hybrids then at Bussey. In Cytological and Genetic Studies of Fragaria (1926) he confirmed Longley's chromosome counts, and to the species with known chromosome constitution, he added the diploid viridis of Europe and the octoploid ovalis of the United States. He also examined eleven cultivated varieties; all were octoploid. In one cross of two diploid forms of vesca (bracteata x rosea) a tetraploid was obtained with larger, thicker, and more crenate leaves than the parents. It was vigorous and fertile. Doubling of chromosomes was noted in seven pollen grains.
In Studies on the Genetics of Fragaria, Ichijima (1930) reported that crosses of nilgerrensis with other diploid species were dwarfs except for three doubtful hybrids. Crosses of the musk strawberry, moschata (2n = 42) x diploids gave seedlings with the moschata chromosome number. One cross of a diploid x octoploid had the chromosome number of the pollen parent but was like the diploid in appearance. Three different forms, vesca rosea, vesca alba, and nilgerrensis had doubled chromosome number in pollen. Moschata had two kinds of chromosomes; two sets with dumbbell shape and one set with spherical shape. Virginiana, glauca, chiloensis, and Chesapeake also had genomes with the two shapes of chromosomes.
A.J. Mandelsdorf and E.M.East (Studies on the Genetics of Fragaria, 1927) made numerous crosses between diploids and concluded that they were entirely interfertile; however, the diploids under study were merely varieties of vesca. In crosses of pink-flowered x white-flowered, pink was dominant. In the F2 they obtained 128 pink to 46 white, close to a 3 to 1 ratio. No crosses were obtained when moschata (2n = 42) was pollinated by a diploid species, but when the cross was a diploid x moschata seed was set. It gave poor germination (0.7 percent) and the few seedlings obtained died within two weeks. When octoploids were used as females, with diploids as males, none set, but the reciprocal was obtained without difficulty. Germination was poor; maternals were obtained; dwarfs did not flower; and all true hybrids were sterile. When the octoploid was crossed with the hexaploid moschata, a good set and a germination of 90 percent was obtained with vigorous, completely sterile plants and dominance of moschata characters. Maternal seedlings, both staminate and pistillate, were also obtained. Mangelsdorf and East, however, were not certain that pseudogamy was the cause.
In Notes on the Somatic Chromosomes of the Seven-Chromosome Group of Fragaria (1929), S.H. Yarnell found that the chromosomes he studied were all very similar in appearance. The only constant difference common to them was that of length; their lengths in microns averaging: 1.7, 1.5, 1.4, 1.3, 1.2, 1.0, and 0.9 for the seven. Usually it was impossible to distinguish differences except between the two shorter and the two longer ones.
Yarnell, in Genetic and Cytological Studies on Fragaria (1931), separated the diploids into four groups: (1) vesca and its varieties, bracteata, californica, americana, rosea, and mexicana; (2) viridis (= collina); (3) nilgerrensis, and (4) a vesca type from China. Viridis gave seedlings that were vigorous in crosses with the other diploids, but were only partially fertile. Nilgerrensis gave only dwarf hybrids in all crosses except those with viridis, and even these hybrids gave no flowers. No true hybrids were obtained in crosses of moschata with diploid species, but some maternals were obtained. The octoploids virginiana and chiloensis set no fruit when used as female parents with vesca, but, used as pollen parents on vesca, they gave both maternal and partially fertile plants. Chiloensis x vesca var.bracteata gave three sterile seedlings. Virginiana and chiloensis both gave vigorous sterile hybrids with moschata pollen, but no set was obtained when they were used as pollen parents. All types of crosses gave at least some completely fertile plants which were like the mother. Increased pairing of chromosomes occurred with increased temperature.
Yarnell (1930) also reported on studies of the tetraploid obtained from crossing two diploids, bracteata x rosea. The tetraploid was selfed and produced tetraploids. These were crossed with the diploids vesca, viridis (= collina), bracteata, and another diploid. All seedlings were triploids.
In 1919 Prof. E. Bauer proposed that Dr. Elizabeth Schiemann (Fig. 7-4) ask for the strawberry collection of Prof. Solms-Laubach of Strasburg, Alsace, so that she could study further Millardet's "false hybrids." In 1930 and 1931 Dr. Schiemann demonstrated that Mangelsdorf and East's explanation for Millardet's patroclinous hybrids (by dominance of the higher polyploid parent) could not be generalized, because hybrids of F. x ananassa (8x) x hagenbachiana (2x) (the frequently found hybrid of vesca x viridis) gave some "diploid" types too, although all with 2n = 35 chromosomes. Another possible explanation for some of Millardet's "false hybrids" was given by Dr. Schiemann after she was able to show that the tetraploid hybrids of moschata (6x) x viridis (2x) could not be distinguished from moschata. These tetraploid hybrids gave fully fertile and more or less constant offspring.
From experiments published in 1931, Dr. Schiemann concluded that females and males (including hermaphrodites) were the two main groups of sex differentiation in Fragaria and, according to Kuhn (1930), suggested that the difference betwen maleness and hermaphroditism was organized by a different allel of the sex factor. Besides the sex factors, she proposed a number of factors for sterility concerning female and male organs as well. Sex changes in plants were often observed. Mutant changes in sex also were found by her. The development of sex organs of the strains used and the different progenies of Schiemann's experiments were investigated cytologically by Rudloff (1930).
From 1933 on Dr. Schiemann (1937) studied hybridization of both the diploid species vesca andviridis x the hexaploid moschata. Similar to Lilienfeld's (1934, 1936) hybrid of moschata x nipponica (elnipponica), Dr. Schiemann succeeded in getting hybrids of moschata x viridis. These hybrids resembled in most characters moschata and in F2 bred more or less true. The autohexaploidy of moschata, demonstrated by Lilienfeld, was confirmed by Schiemann's results, as was also homology between the genomes of viridis and moschata.
Homology of the genomes of moschata with those of vesca was concluded by Schiemann (1944) from a tetraploid hybrid (vesca x moschata) which produced normal pollen grains and had always fourteen bivalents in meiosis I. Thus, the homology of the three diploid species vesca, viridis, and nipponica could be established.
After 1945, Schiemann's associate, U. Nurnberg-Kruger, took over the hybrid material vesca x nilgerrensis, and G. Staudt started investigations on taxonomy and geographical distribution of the genus Fragaria. The need for an exact knowledge of taxonomy for the planned investigations on the phylogeny of the genus resulted in an extensive collection of strawberries from all over the world. Dr. Schiemann's experiments, in which orientalis-like plants originated in the F2 of the pentaploid hybrid of vesca x moschata, initiated G. Staudt's investigations on the phylogenetic relationship and sex differentiation of the tetraploid orientalis.
Open questions in the genus Fragaria was published in 1951 in which Dr. Schiemann pointed out the main lines for further research: analysis of the diploid species in Europe, of the genome relationship between the European and Asiatic species, and analysis of the tetraploid orientalis, as the first steps in a study of the evolution of the polyploid species.
Schiemann's latest paper (1958) dealt with a subfertile hybrid which had originated in 1923 from F. x ananassa x virginiana. Besides having characteristic teratological malformation of the leaves, the plants, propagated vegetatively, came to flower for the first time after sixteen years. From that year on, the malformation of the inflorescences and flowers decreased.
G.Staudt, a student of Dr. Schiemann, in Cytotaxonomy and Phylogenetic Relationships in the Genus Fragaria (1959) concluded that American octoploids were not derived from American vesca types, but may have come from East Asia and may have been derived from East Asian vesca types. He suggests that the octoploids virginiana and chiloensis were derived from different ancestors. In a later paper, The Origin and History of the Large-fruited Garden Strawberry in Germany, Staudt traced the introduction of virginiana and chiloensis into Europe. He noted that the chiloensis plants "have no winter hardiness" and that in many places no flowers developed. The Mapuche or Huilliche Indians were the first to cultivate the strawberry in Chile. By 1759 x ananassa, the hybrid between virginiana and chiloensis was known. Staudt's Taxonomic Studies in the Genus Fragaria (1962) is a detailed study of Fragaria of America and Europe. Six species are recognized: vesca (with 4 subspecies), viridis, moschata, virginiana (4 subspecies), chiloensis (4 subspecies), and x ananassa. Included in vesca are subspecies americana, bracteata and California. Subspecies glauca, platypetala, and grayana are given for virginiana and subspecies lucida, Pacifica and sandwicensis for chiloensis.1
Under the leadership of Dr. S.C. Harland (Fig. 7-5) and assisted by Miss Edna King, Miss M. Smith, Miss Thelma Smith and A.S. Islam, J.R. Ellis, and J.K. Jones a major genetical research project on the strawberry was undertaken along seven lines.
(1) Mildew resistance. Harland and Miss King (1957). Mildew susceptibility was found to be due to two dominant genes in the diploid F. vesca. Both genes must be absent to obtain a resistant plant. In the F2 of resistant x susceptible a 15 to 1 ratio of susceptible to resistant was obtained. Cytoplasmic effects were noted: if the resistant plant was used as a female the F1 was less susceptible than in the reciprocal. When accidentally a haploid with only seven chromosomes was obtained, it was immune to mildew, but when it was doubled to a diploid form again, it was partly susceptible. When redoubled to become a tetraploid, it was still more susceptible. The most resistant of all octoploids tested was a chiloensis from the Oregon coast.
(2) Studies in ploidy, J.R. Ellis, Thesis. (a) Commercial possibilities of decaploids. Decaploids had been produced by Darrow and by Scott, by obtaining 4x vesca, crossing it with cultivated octoploids to obtain hexaploids, then crossing the hexaploids with octoploids and selecting out the ten-ploids. Ten-ploids were then intercrossed and Scott raised several hundred seedlings from which commercial types could be selected. Other methods of obtaining decaploids were studied by Ellis, such as crossing 2x x 8x and doubling the resulting 5x to a 10x, or crossing the 4x vesca with 16x virginiana and with 16x cultivated varieties (obtained by doubling the chromosome number of virginiana and of varieties). The work had not progressed far enough to show the most promising method.
(b) Octoploids with some vesca chromosomes. Two methods were planned for obtaining octoploids from vesca and moschata -- (1) doubling of chromosomes of vesca (2x) x moschata (6x) hybrids; (2) doubling of chromosomes of 4x vesca x moschata (6x) hybrids and then back-crossing to moschata. The first method was found to be difficult and only one hybrid was produced, a very vigorous octoploid. The second method was therefore followed. The cross 4x vesca and moschata (6x) gave only vigorous pentaploids, all fertile males. Ellis then theorized that moschata could have been derived from the allopolyploid of 4x vesca x nubicola (2x) as Staudt had suggested. Moschata would then contain 4 vesca genomes plus 2 others. The cross 4x (vesca) x 8x variety was also used to obtain 6x; these were crossed with vesca (2x), and then doubled to 8x. Two methods of obtaining octoploids from vesca and cultivated varieties were successful (see p. 109).
To obtain other kinds of octoploids, two methods were successful: decaploids were crossed with hybrid hexaploids, and hybrid tetraploids were doubled. The first kind of octoploid was highly sterile and the second octoploid types have not grown large enough to estimate their value. As Vilmorin (1898) and many others had suggested, Ellis concluded that it was unlikely that present octoploids had vesca in their immediate ancestry.
Ellis (1958) reported finding functional double unreduced gametes andsuggests this as a possible way for moschata and virginiana and chiloensis to have originated, with no intermediate polyploid types, from diploids native to their habitat. Moschata would be derived from a double unreduced x, a single unreduced gamete, to give 4x x 2x = 6x and the octoploids from the union of two double unreduced gametes. He suggests the probable value of an octoploid with vesca, moschata, and x ananassa parentage.
(3) Intergeneric crossing. Ellis (1962) reviewed crosses between F. vesca and Potentilla made by others. Ellis obtained hybrids from the Sans Rivale strawberry x Potentilla fruiticosa and x P. palustris. Surviving seedlings were 5x and 6x. They were both female and male sterile. 4x vesca x P. fruiticosa seedlings were 3x. Octoploids x P. palustris (6x) were 7x with the most vigorous resembling Fragaria. Their rose-pink flowers were male sterile and slightly female fertile.
(4) Induction of haploids. No haploids have thus far been obtained. Islam (1961) studied the accidental haploid plant obtained from a vesca x x ananassa and compared it with the parent vesca; it was smaller in all its parts.
(5) Genetics of perpetual flowering on the octoploid level. Harland and Dr. Thelma Smith and Miss Margaret Smith, from a very limited crossing, obtained a possible 9 perpetual-flowering to 7 normal ratio with the perpetual being AB and the normal ab. Ellis then proposed that perpetuals x normal selfed would give 9 AB:3 Ab: 3 aB to 1 ab; the perpetual-flowering resulting from complementary genes AB.
(6) Propagation methods. Harland and Margaret Smith made a study of rapid propagation by using small one-inch long pieces of roots and small runner tips. The one-inch pieces of roots developed buds and made plants very slowly.
(7) Selfing. Through selfing some varieties lost little vigor, others had very weak progeny.
Effects of induced polyploidy: Islam, Jones, and Ellis found the constant characters associated with chromosome doubling in strawberry to be:
(1) Thicker leaves with larger serrations;
(2) Reduction in the length to breadth ratio of leaflets; that is, rounder leaflets;
(3) Increase in width of the basal angle of the terminal leaflet.
Islam in Possible Role of Unreduced Gametes in the Origin of Polyploid Fragaria (1960) reviewed the pertinent literature, and from this and his own studies concluded that "the frequent production of unreduced gametes may have been the principal method of origin of polyploids in Fragaria." Their origin may be either the union of gametes with the full chromosome number of one parent and the reduced number of the other, or with the full number of both parents, called double non-reduction. As to the origin of tetraploids, Islam pointed out that Schiemann (1951) had noted 4x among the F1 progeny of the cross 2x viridis (= collina) x 2x vesca. Accordingly, Islam suggested that the tetraploids orientalis and moupinensis may have originated through hybridization between the two diploids with overlapping habitats followed by chromosome doubling, or by the union of an unreduced with a double unreduced from two diploids. He suggests the latter method as the origin of hexaploids, for Lilienfeld (1933) obtained a hexaploid from the progeny of the 4x hybrids between the 6x F. moschata (elatior) and the 2x vesca (nipponica). Any of several diploid species may have been the progenitor of the hexaploid species. Octoploids could have resulted by non-reduction from undiscovered 4x or 6x (in crosses with diploids) American species, or by nonreduction from the Asiatic orientalis or moupinensis and subsequent migration of the 8x to America. The first seemed most plausible.
Among other articles concerned with cytogenetical research with strawberries is A Contribution to the Question of Genome Relations in Some Species of Fragaria, by N.A. Dogadkina, which reports that the cross vesca x x ananassa resulted in few seedlings, only four of which flowered. Chromosome behavior proved the homology of the genomes of vesca with one of those of x ananassa. The cross orientalis (2n = 28) x elatior (moschata) (2n = 42) was first made by Fedorova in 1932-33 in both directions. All the hybrids were pentaploid and twenty-seven plants were at least somewhat fertile. It was concluded that the two genomes of orientalis were homologous with two of moschata (elatior), that all three of moschata were homologous, that is from closely related species, but that moschata is not a true autohexaploid. Dogadkina points out that crosses made by Kihara (1930) and Fedorova (1934) indicate that three of the genomes of x ananassa are homologous with those of moschata. Fedorova obtained a tetraploid from vesca x moschata that indicated homology of the genomes of vesca with those of moschata, so that at least three of the x ananassa genomes seem closely related to that of vesca. Fedorova suggests that orientalis is a link between vesca and moschata.
Cultivated large-fruited strawberries and the species Fragaria chiloensis, virginiana, and ovalis from which they come, intercross freely and their hybrids, with certain exceptions, produce fertile seedlings. But the cultivated varieties and the octoploid species from which they come do not readily cross with diploid, tetraploid, and hexaploid species; seedlings of such crosses are nearly or entirely sterile. However, this does not mean that the latter species should be dismissed from consideration as breeding material. These other species need to be surveyed for qualities that might be useful in cultivated varieties. Then, methods of transferring desired qualities to cultivated varieties need to be found. Attempts were made as long ago as 1760 by Duchesne to make some of such crosses, but even today we do not know for certain of any Successful attempts-success, that is, in terms of transferring desirable qualities from these other species to large-fruited commercial sorts. No progress was made along these lines until the hybrids were made by Dermen and Darrow in 1944, and later improved by Scott (Scott, 1951). More recently Lebedov in 1957 reported obtaining a hybrid of a cultivated variety x moschata variety that produced well and Katinskaja in 1963 reported obtaining useful sorts from cultivated varieties by vesca, by viridis, and by moschata(see under Russia).
The barrier to hybridization in large part has been differences in chromosome number. There are now four natural species groups known, and the phenotypic classifications of strawberry species by botanists constitute classes within these four groups. All such phenotypic classifications based on gross morphology are placed under the four chromosome number groups-2x, 4x, 6x, and 8x. There are in all about eleven strawberry species-classifications which seem to rest upon important enough differences to be valid, although more than 45 have been described. These eleven belong to the botanical and chromosome groups as follows:
|Species and Native Country||Geneticist Reporting|
|Diploid: (14)||1. F. daltoniana||-Asia||Darrow, 1937|
|"||2. F. nilgerrensis||-South Asia||Shiemann, 1951|
|"||3. F. nubicola||-" "||Staudt, 1959|
|"||4. F. vesca||
-Circumpolar, north Africa,
mountains of South America
|"||5. F. viridis (=collina)||-Central Europe||Ichijima, 1926|
|Tetraploid: (28)||6. F. moupinensis||-East Central Asia||Staudt, 1951|
|"||7. F. occidentalis||-" " "||Petrov, 1934|
|Hexaploid: (42)||8. F. moschata (=elatior)||-Central Europe||Longley, 1928|
|Octoplod: (56)||F. chiloensis||
-Coast Alaska to central California,
south Chile, mountains
10. F. ovalis (=cuneifolia,
|-Western North America||Ichijima, 1926|
|"||F. virginiana||-Eastern North America||Longley, 1926|
Some diploid species do intercross readily, but most hybrids have reduced fertility. The two tetraploids are said to be interfertile and their hybrids fully fertile. The three octoploids are entirely interfertile.
F. daltoniana is a little known, small, one-flowered species of South Asia. Its chromosome number is reported as diploid; botanically it belongs in the first group and its response as a parent places it there. The other four species in the diploid group (2n = 14) have in general small, usually thin-leaved plants which bear smaller berries than plants with higher chromosome numbers. Nilgerrensis is a many-seeded, tasteless, or ill-flavored species with white flattened berries on erect pedicels. Though vesca x nilgerrensis are usually dwarfs, Nurnberg-Kruger (1958) reported normal seedlings. Nubicola is similar in gross appearance to vesca, but hybrids with vesca are usually sterile. Vesca is of interest for its wide distribution, its aromatic, high-flavored berries, and its everbearing forms. Viridis has hard to cap, good-flavored, firm fruit. Though known for three hundred years, it has so far furnished no good garden varieties.
A single vigorous sterile triploid was found in eastern Oregon in 1949. It was collected by R.C. Rosenstiel and the chromosome count made by H. Dermen (unpublished). It resembled 4x vesca and probably resulted from the union of an unreduced with a reduced gamete of a diploid vesca. Triploids have been obtained by Dermen and Darrow from 2x x 4x crosses (Dermen and Darrow, 1938, Fig. 7-6).
Moupinensis and orientalis are hardy species of West and North China, Korea, Manchuria and Siberia. They undoubtedly have arisen from diploids, probably from crosses of two diploid species. Spontaneous tetraploids are known to arise from crosses of the diploids viridis x vesca and they resemble orientalis. Tetraploids have been produced from vesca (1) by using colchicine, (2) by crossing vesca (bracteata) x vesca (rosea) (Yarnell, 1929), (3) by crossing vesca x moschata (Mangelsdorf, 1927), (4) by crossing moschata x vesca (nipponica) (Lilienfeld, 1933, 1936), (Schiemann, 1951), and (5) by crossing vesca x virginiana (East, 1933, 1934).
Very vigorous sterile pentaploids (chiloensis x vesca) (Bringhurst, 1964) have been found in the wild on the coast of California. Pentaploids from vesca x octoploids have been produced by Fedorova (1932), Dermen and Darrow (1938, Fig. 7-7), Scott (1951), Yarnell (1931a), Islam (1954), Ellis (1958), and by Mangelsdorf and East (1927). The latter also obtained pentaploids from vesca (both americana alba and rosea) x ovalis (glauca) and from vesca (bracteata) x virginiana. None that flowered was fertile. Usually the reciprocal cross of the octoploid x diploid fails entirely. However, in an attempt to produce apomictic seedlings of the Fairland and Temple varieties by pollinating with vesca, Darrow (unpublished) obtained a very large number of seedlings, all of which were vigorous true hybrids and set at least a few seeds. Only one had a fair development of fruit. Ellis obtained pentaploids from crosses of vesca x moschata, all very dwarf; from tetraploid vesca x moschata, all vigorous and all males; from vesca (2x) x chiloensis, virginiana, and X ananassa (Ellis, 1958).
F. moschata (Plate 7-1), the one hexaploid, is cultivated slightly in the gardens of Europe because of its vigorous plants and very vinous-flavored fruit. It may have originated by natural crosses of vesca x viridis or vesca x nubicola (Staudt, 1959), unreduced pollen or eggs of one parent furnishing the two extra sets of chromosomes, followed by doubling of the resulting triploids. Hexaploids have been produced by crossing cultivated varieties with 4x vesca (Scott, 1951) (34 seedlings) and by the reciprocal 4x vesca x octoploids virginiana, chiloensis, and cultivated varieties (Dermen, 1938-1939, unpublished) (Figs. 7-8, 7-9), (Ellis thesis, 1958), (Islam thesis. 1954), (Kluge, 1959).
These have been obtained by Scott and Darrow from (F. x ananassa x 4x vesca) x x ananassa and, although one was selected for its fertility, it was finally discarded (unpublished). Ellis (1958) obtained seven-ploids from the 8x cultivated varieties x 7x hybrids and the reciprocal.
Most cultivated varieties are octoploids and descendants of chiloensis and virginiana. A few varieties have F. ovalis in their ancestry also. The octoploids are native to the Americas and the Hawaiian Islands. Together they have the characters that by intercrossing have provided the basis for the present large strawberry industry. Dermen obtained an octoploid by doubling and redoubling an Alpine vesca by using colchicine (Fig. 7-11).
To synthesize octoploids from lower chromosomal species, Ellis has suggested crossing hexaploids with pentaploids and looking for 8x seedlings from a union of hexaploid pollen, or egg, with unreduced pentaploid pollen, or egg. Other ways he suggests are:
2x x 6x, looking for non-reduction in both parents.
2x x 6x, looking for normal reduction division, but the 4x doubling in the embryo or by colchicine (Plate 7-2b).
2x x 6x, looking for normal reduction division to obtain a 4x, followed by unreduced egg and pollen uniting.
2x x 8x -----> 5x (unreduced) x 2x -----> 6x x 5x (unreduced)---->8x.
2x x 8x -----> 5x (unreduced) x 2x -----> 6x x 2x ------>4x (unreduced) and doubled ------->8x.
2x x 8x -----> 5x (unreduced and doubled) x 8x ----->14x x 2x --------> 8x.
Ellis obtained octoploids from 6x (4x vesca x F. x ananassa) x 10x (4x
vesca x 16x of cultivated). These octoploids were mostly vigorous and highly sterile, completely female sterile, producing anthers with 3 per cent good pollen. Another octoploid was obtained from 2x vesca x hybrid hexaploid and the seed treated with colchicine to double the chromosome number. Fertility of the plant was not given. One octoploid was obtained from colchicine treatment of seed from vesca x moschata but this cross was not promising. Fedorova (1946) obtained octoploids, some of which were fully fertile, from selfing a partially fertile 7x seedling of F. x ananassa x moschata. Dermen (1938, unpublished) obtained a fertile octoploid by crossing Dorsett with an 8x vesca obtained by colchicine treatment of 4x vesca (Fig. 7-12).
No naturally occurring plants with chromosome numbers higher than fifty-six have been found. Scott obtained thirty-five plants all with sixty-three chromosomes (nine-ploid) from a ten-ploid x eight-ploid cross (1951). They were vigorous, but most were only slightly fertile. Another similar cross produced eighteen seedlings which were likewise only slightly fertile and he concluded that nine-ploid seedlings would be of little value. Fedorova (1946) and Ellis (1958) also obtained nine-ploids.
Decaploids have been obtained by several methods:
Dermen (1938, unpublished) from the cross 2x vesca x 8x Dorsett variety obtained fully fertile 10x seedlings by colchicine treatment of seeds. He also obtained a decaploid by doubling the chromosome number of a seedling of a vesca x Dorsett (Fig. 7-13). Ellis (1958) reported obtaining five ten-ploid seedlings by treating with colchicine 2x vesca x virginiana and x chiloensis, and twelve twelve-ploid seedlings following treatment of 4x vesca crossed with the same species. The vesca x virginiana decaploids were highly fertile. In 1942 at Beltsville, Md., nearly sterile hexaploid seedlings (Fig. 7-8) (8x pistillate variety x 4x vesca) were crossed back to cultivated varieties (Fig. 7-9) and over seven hundred and fifty seedlings raised. About two hundred flowered in 1943. Seven of the more fertile were saved, of which Scott found one partially fertile with 2x = 49 (seven-ploid) and six fully fertile with 2n = 70 (decaploid) chromosomes. Most of the seedlings had spongy fruit like vesca, with large air spaces, and only six were selected for their large, firm fruit.
Scott (1951) repeated the cross octoploid variety (Midland) x 4x vesca and obtained eighteen probable hexaploids along with eight counted ones and eight octoploids; from a cross of an octoploid with two of these hexaploids he obtained thirty probable and twenty-three counted seven-ploids, eight octoploids, six decaploids, and one eleven-ploid; and from a decaploid x decaploid cross he obtained fifty-one probable and twenty counted decaploids. Decaploids x octoploids gave twenty-six probable and nine counted nine-ploids. Most of his hexaploids were very vigorous and failed to blossom or were sterile. He assumed that the ten-ploids originated as six unreduced gametes of the hexaploid uniting with the four reduced number of the octoploid (28) to make ten-ploid. Though he assumed that the fully fertile octoploid seedlings of his crosses resulted from self contamination, they may have resulted by apomixsis due to failure of the first reduction division of the egg-mother-cell.
Most interestingly, Scott, out of fifty-two fully fertile decaploid seedlings, obtained twenty-nine seedlings with a high per cent of good pollen. The plants were indistinguishable in foliage and growth habit from cultivated varieties. Although the fruit of most decaploid plants was spongy, all final selections had firm fruit with the high aroma of vesca; the fruit was not as large as that of the newer cultivated varieties. These decaploids had fourteen chromosomes from vesca and fifty-six from the cultivated octoploid. Fedorova (1946) also reported fertile decaploids.
It seems obvious, therefore, that it is entirely possible to breed decaploid strawberries with the high aroma of vesca. Breeding can be continued on the decaploid level by the continued combining of 2 genomes of vesca with 8 of the cultivated, while using different octoploid ancestors to increase the array of decaploids (Plate 7-3a).
Seedlings of cultivated octoploid varieties and of virginiana and chiloensis have had their chromosomes doubled to sixteen-ploids (Dermen and Darrow, 1938, Ellis, 1958). The 16-ploids of Ellis were hybridized with 4x vesca to obtain decaploids (see above). From colchicine treatment of seedlings of varieties crossed with chiloensis, Hull (1960) reported obtaining eighteen chimeral sixteen-ploid seedlings whose second apical cell layer was affected in sectors and nine seedlings where the second layer was affected in the entire crown. In a comparison of plants having sixteen-ploid at least in the second apical layer, with plants of the same seedlings having eight-ploid in the sec ond layer, the sixteen-ploid had larger fruit, indicating a possible advantage of breeding at the sixteen-ploid level. Ellis (1958) obtained four sixteen-ploid selfed seedlings of chiloensis, five of virginiana, and three of cultivated varieties. The sixteen-ploid selfed chiloensis seedlings were very vigorous while those of selfed virginiana and of octoploid varieties were slower in growth. Seedlings of sixteen-ploid Huxley seedlings were, however, comparable in vigor to the normal Ettersburg 80 (Huxley). Dermen's (1938 unpublished) sixteen-ploid was from treating Dorsett seedlings with colchicine (Fig. 7-15).
Scott obtained one seedling with seventy-seven (11 genomes) chromosomes -- an unreduced octoploid uniting with three sets from a reduced hexaploid. The seedling was only partially fertile and was less vigorous than the decaploids. Fedorova (1934) reported 7-, 8-, 9-,11-, 12- and 14-ploids from F. x ananassa x moschata and 13x from the 7x crossed back to moschata. Ellis obtained three 11-ploid from a 6x hybrid x a 10x hybrid (vesca x F. x ananassa). Scott failed to find any aneuploids, but Ellis obtained odd chromosome numbered seedlings from the cross 4x vesca x 16x virginiana, most of which were stunted. He obtained 15-ploids from 5x x 10x (vesca x F. x ananassa), 18-ploids from 5x (unreduced) x 16x x F. x ananassa. Islam (1954) reported 12x, 13x, and 14x seedlings from open-pollinated F1 of vesca x Bradley Cross (8x). Fedorova (1946) reported obtaining 12-ploid and 14-ploid from the selfed seedlings of F. x ananassa x moschata. Dermen (1939, unpublished) obtained a sterile 12-ploid (4x vesca x Dorsett) (Fig. 7-14), a fertile 32-ploid (of a 16-ploid Dorsett seedling) by colchicine treatment (Fig. 7-16).
Chromosome doubling of germinating seedlings, especially of diploids, has been quite successful. Dermen and Darrow (1938) obtained tetraploid vesca by treating germinating seedlings with 0.2 per cent colchicine solution for twenty-four hours. A sixteen-ploid seedling of Dorsett was obtained by treating seedlings for five and six hours. Hull's sixteen-ploids resulting from colchicine treatment of seeds were chimeral. Ellis reported (1958) that 78 per cent of the surviving treated seedlings of vesca were tetraploid. He used 1 percent colchicine for twenty-four hours.
Ellis (1958) concluded that hybridization between the induced polyploid forms of vesca and cultivated varieties was a very favorable method for producing ten-ploids. Though Ellis pointed out that the ten-ploids of both Scott and himself produced smaller fruit than that of the present octoploids, this does not seem to warrant any conclusion as to the possible size of fruit borne by decaploids. The size of present cultivated varieties is the result of selection of the larger-fruited sorts from millions of seedlings and, given as long and as extensive a program, just as large-fruited decaploids with vesca aroma might be obtained as easily.
It would also seem entirely possible to obtain octoploids with the strong aroma of vesca, or of moschata, so that, by their use as intermediaries, stronger aroma could be transferred to cultivated varieties. This would seem more logical than the origination of separate groups of decaploid and octoploid varieties indistinguishable in appearance. Results so far indicate that decaploid x octoploid crosses would be sterile or nearly sterile, so that breeding would have to be continued within the separate groups. Far less confusion would result if all varieties were decaploid or octoploid.
* By Dr. G. Staudt
1. Having grown hundreds of collections of virginiana and chiloensis, I prefer the botanical classification given on pp. 122-123, as being most realistic. No subspecies are kept, for the differences within the species are great in most areas. George M. Darrow.