The more closely one observes individuals, be they atoms, molecules, algae, bees, flowers or man, the clearer becomes their uniqueness. Thus, though for purposes of classification one may group roses into various color classes, no two seedlings are really exactly alike. External environment factors modify the expression of the genetic factors involved. High day temperatures lighten flower color. In the fall, continuing warm nights and relatively cool days allow the margins of the petals to develop very beautiful color accents due to various modifying factors. The over-all color intensity is deepened. In spite of these fluctuations, certain definite color groupings may be made. Since a number of popular varieties such as Super Star (Com. syn. Tropicana), Hawaii, San Francisco and Tickled Pink are in the scarlet-vermilion to signal red color range, a discussion of the genetic basis of these colors may be helpful to rose breeders.
Ann Wylie (1954) reported that pelargonidin is responsible for the brilliant hues found in Paul Crampel, Gloria Mundi and Gloire du Midi. These colors were unknown in garden roses before 1930. Whether this mutation from cyanidin to pelargonidin occurred more than once cannot be determined since we do not know the origin of all orange-flowered dwarf polyanthas. The occurrence of pelargonidin in Sondermeldung (com. syn. Independence) had been proven by paper chromatography studies. The normal red color pigment cyanidin is also present. Whether the color of Independence is due to an independent mutation from that responsible for the Gloria Mundi group is not known. It must have occurred as far back as Eva which occasionally produced orange-cinnabar seedlings. Such a mutation is technically known as a recessive one since the normal red color predominates. Red roses have the dominant factor for cyanidin pigment formation in at least two of the four chromosomes. Most modern garden roses are tetraploid, that is, have seven sets of four homologous chromosomes making a total of 28. Garden roses such as Chrysler Imperial, "carry" the pelargonidin factor in one of the four chromosomes even though red in color. Accordingly a small percentage of scarlet-vermilion color is obtained when Chrysler Imperial is crossed with Tropicana. [CybeRose note: In his 1960 paper on the Inheritance of Magenta, he assumed that Chrysler Imperial is MMmm rather than MMMm.]
My own interest in these colors resulted from a desire to get a larger hybrid tea flower having the Independence color. Accordingly, I crossed Dean Collins (British Horticultural Color Council claret rose) with Independence. The color range displayed in the resulting seedlings was truly amazing. Since, in making this cross and the rest reported, my main objective was the production of either non-fading red roses or ones in the scarlet-vermilion-signal red range, these are arbitrarily called "good" colors. All of the seedlings having the magenta or M factor (1) were classified as undesirable in color. The resulting segregation is given in Table I.
|A. "Good" Colors Arranged According to Major Color Groups|
|1. Rose Red - Currant Red Series|
|2. Signal Reds|
|4. Camellia Rose Series|
|5. French Rose Blends|
|Total Good Colors||59||230||160||25||63||69|
|B. "Bad" Colors|
|1. Solferino Purples Series|
|2. Tyrian Purples Series|
|3. Phlox Pink Series|
|4. Delft Rose Series|
|5. Tyrian Rose Series|
|Total Undesirable Colors||60||61||53||12||27||19|
About one-half of the seedlings evidently received an M factor from Dean Collins and so varied in color from phlox pink, having a strong magenta hue, to rose pink, a very light pink having only a slight magenta hue. Twenty-six of the seedlings were in the signal red series, very attractive colors free of any magenta hue and easily visible red cyanidin pigment. One of these was introduced as San Francisco, a fluorescent signal red. The dark reds, blood, Orient, Turkey and cherry were unexpected. Their occurrence in populations from all crosses involving hybrids which in appearance vary from light scarlet to signal red indicates that the factor for red cyanidin pigment cannot produce enough red color to be readily noticeable when present in only one of the four chromosomes involved.
The cross of Queen Elizabeth x San Francisco gave a remarkable range of colors. Particularly of interest is the reduction of "undesirable" colors from 50 per cent to about 25 per cent in this cross and the other four analytical ones reported. Analysis of populations from many crosses involving Queen Elizabeth indicate that it carries only one dominant M or magenta factor. This does not cause an undesirable hue in Queen Elizabeth because the basic A factor for anthocyanin pigment is not present (2). Populations obtained by self-pollinating Queen Elizabeth do not show any red flowered plants.
Only 1/2 of the pollen of San Francisco carries the basic A factor for cyanidin formation. In cytogenetic terminology this means that only one of the four chromosomes involved carries the factor for cyanidin and the other three carry the pelargonidin (geranium red) factor. Since only 1/2 of the egg cells of Queen Elizabeth carry tbe M factor and 1/2 of the pollen of San Francisco the A factor, 1/2 x 1/2 or 1/4 of the plants will have an undesirable color, that is, carry both M and A. The actual number of seedlings, 61, is well within the range of sampling error in its deviation from the expected 1/4 of 291 or 72.
H55060/1 is a light scarlet hybrid obtained by crossing Queen Elizabeth with Independence. It also gave only about 1/2 of the plants having the undesirable magenta color when backcrossed to Queen Elizabeth. 51043/52 is similar to San Francisco in color but not as brilliant in intensity. It was obtained by selfing Dean Collins. Here, slightly more than 1/4 of the plants were tinged with magenta. Finally, H55016/663 is a sister seedling of San Francisco but Dutch vermilion in color. Here again, slightly more than 1/4, i.e., 27 out of 90, were undesirable in their magenta color.
The range in hue of the various color classes was greater even than Table I adequately conveys. Thus, the scarlet-vermilion series is of special interest because of the unusually lovely colors in this group of plants. The breakdown of this color group is given in Table II for each of the crosses reported. Careful study of these colors as well as those in the camellia rose series makes it seem probable than an allelomorphic series of factors for pelargonidin formation is involved. Possibly a number of recessive mutations inhibiting different phases of the reaction system involved in formation of the normal cyanidin or red pigment occurred in the history of the rose. Actually the various polyanthas in which the mutation to pelargonidin occurred differ from one another considerably in the pelargonidin or geranium red hue expressed. Independence may indeed be the result of a mutation which interferes with the cyanidin pigment formation much later in the stage of enzymatic action than the lighter scarlet mutations. Possibly some day research into the DNA-RNA reaction system as regards color pigment formation in roses may shed light on this problem.
|Poppy Red||Fire Red||Nasturtium
|Queen Elizabeth x San Francisco||14||8||1||1||-||-|
|H55060/1 x Queen Elizabeth||29||11||1||3||1||1|
|San Francisco x H55016/663||15||1||-||-||1||-|
|H55060/1 x San Francisco||9||4||3||1||-||-|
Another consideration in the breeding of roses with these lovely colors is that this mutation for some unknown reason also is associated with factors for weaker plant growth. Whether the same gene or genetic factor causes both is not clear. However, the evidence from many crosses indicates that when one of the parents involved carries the pelargonidin factor, the resulting hybrid population is weaker than average. Thus, Queen Elizabeth x San Francisco averaged 41 per cent strong, vigorous plants as compared to crosses with Red Delight, 69 per cent; Pearl of Aalsmeer, 57 per cent; H56024/20, 66 per cent; and H56024/39, both reds 87 percent. Intercrosses of H55060/1 x San Francisco were from one to two feet shorter and generally much smaller than hybrids between cyanidin colored varieties. Tropicana, Baccara and Aztec gave similar results. Since these come from quite unrelated ancestry, the results can hardly be due to inbreeding.
The above reaction is, of course, in line with the well-known fact that mutations generally are harmful. In order to counteract this tendency of pelargonidin colored hybrids to give weaker seedlings, advantage may be taken of the unusual vigor of Queen Elizabeth. Though only 41 per cent of the plants in the population obtained by crossing Queen Elizabeth with San Francisco were normal in vigor, among them some had beautiful scarlet to signal red coloration. Again, though Baccara x H5509/183 ([Captain Thomas x Crimson Glory] x Happiness) gave only 46 per cent strong plants, Queen Elizabeth x Baccara gave 82 per cent strong plants. Evidently Queen Elizabeth has factors for vigor which counteract the generally weak growth factors inherited from Baccara.
Though Baccara, San Francisco and Tropicana all tend to transmit factors for weak growth, they each have features making them invaluable in breeding. Baccara transmits remarkable substance to its hybrids; San Francisco, a high degree of rust resistance; and Tropicana, a high degree of mildew resistance. The problem, of course, is to combine these desirable virtues with the vigor factors from Queen Elizabeth. Though complicated, considerable progress is being made.