JARS v42n2 - In Search Of A Yellow Evergreen Azalea

In Search Of A Yellow Evergreen Azalea
Bob Badger, Kent, Washington

Reprinted with permission from The Azalean, September, 1987

In the early 1960's, at a dinner at Ivan's Restaurant in Puyallup, Washington with Augie Kehr, he made a remarkable statement, "I think it is possible to produce a yellow-flowered evergreen azalea." What a shocking statement that was! His theoretical method involved using deciduous azaleas as the source of genes for yellow flower color and the possible creation of tetraploid forms through the use of colchicine. The idea seemed plausible, yet so difficult to achieve.
About eight years ago, in response to an article I wrote in the Seattle Rhododendron Society's newsletter, the Northwest Hybridizers Group was formed to gather amateur hybridizers together to discuss successes, failures, and the complex problems in hybridizing rhododendrons. The group now has a southern section in Tacoma, and at the two monthly meetings, as many as sixty or seventy hybridizers exchange ideas. They have learned much about color inheritance.
Now, I believe it is easily possible to create a yellow evergreen azalea. I have certain strong ideas, so I wrote Augie to ask him of his success to date.

Notes On Hybridizing For a Yellow Evergreen Azalea From Dr. August Kehr - February 1987
What a marvelous memory you have to remember a discussion at a dinner in the 1960's! In response to your letter, I am very actively working on breeding a yellow evergreen azalea. The method being used is exactly the one I proposed in 1963, before I had even made the first cross.
It was somewhat daring for me to write an article on a method that was purely theoretical. You will find the article in the publication that Mr. Leonard Frisbie published called "Rhododendron". The complete reference is: Kehr, A.E. 1963. "A Potential Method to Produce a Yellow-Flowered Evergreen Azalea", Rhododendron 13 (3): 4-5.
In the 24 years following the original article, I have found out many things through experience. Although the method is unchanged, the choice of breeding material has changed drastically. Now for some notes:
1.  My original parents in the yellow azalea program were: R. calendulaceum 'Colossus' as a source of the carotinoid pigments: and tetraploid evergreen azaleas 'Wako', 'Taihai', 'Banka', and 'Getsutoku', which genetically functions as a tetraploid.
2.  Hybrids between R. calendulaceum 'Colossus' and the four azaleas are easily made, the cross produces lots of seed, the seed germinate readily, but the offspring are weak growers. However, at maturity the hybrids are both male and female fertile, and F2's are readily obtained (also weak growers).
3.  The cross succeeds in the direction shown but not the reverse (i.e., 'Taihai' x R. calendulaceum 'Colossus' is unsuccessful). This is exactly the situation found in crosses with the Malesians, which I will report on at Eugene. (Did you know Malesians could be hybridized with scalies and non-scalies outside the Vireyas?).
4.  My problems (unforeseen) were that the white flowered tetraploids used as evergreen pollen parents contained genes for pigment production of anthocyanins - hence the yellow was hidden by the red anthocyanin pigments. I have corrected this by using a recessive white evergreen parent. The above whites are white because of a single gene that prevents all color from showing. The recessive white is white because of the above gene plus the fact that all the anthocyanin genes are also recessive. Likewise, R. calendulaceum 'Colossus' is orange. The color orange comes from anthocyanin pigments (water soluble red) plus the carotenoids (water insoluble yellow). Thus both of my earlier parents carried anthocyanin pigments. For this reason all the hybrids were orange - not yellow. I failed to foresee this pitfall.
5.  I have since been using several evergreen azaleas that are cream-colored because they carry water soluble flavones. These include 'Olga Niblett', Pryor 76-89, Pryor 75-305, R. kaempferi 'Cream', etc. The Pryor lines are the remnant (actually the best lines) of the Beltsville program that fell under my administration as Chief of the Vegetables and Ornamentals Branch. It now appears that Bob Pryor's work could get only cream colors because the breeding material contained only flavones. I maintained two lines as indicated. Pryor 76-89 is a light cream color. It is interesting because some of the flowers lack both pistils and stamens. Even when good pistils are present, they are nonfunctional. The pollen, however, is fully functional. Pryor 75-035 is even more interesting. It is a dwarf with badly crinkled leaves, but is a light yellow the best color of all the material. However, I have never been able to propagate it, and my only plant is slowly degenerating. Pryor 75-305 is fertile, female and male. I suspect the badly crinkled leaves come from chromosomal aberrations resulting from the initial crosses Bob made with deciduous x evergreen parents at the diploid level. I was almost dumfounded to find in some of my progenies from Pryor 75-305 the exact same crinkled leaved segregates, perhaps further evidence of the transmission of the probable chromosomal aberration.
6.  The hybrids and derivations of the parental material described in (5) are all diploids. Try as I might, they do not cross with the yellow-flowered R. calendulaceum plants I have. Hence my efforts last year and now are to double the chromosomes. My basement is full of seedlings treated with colchicine. This year I am trying the acenaphthene method described in my chapter of Fred Galle's Azaleas book. This method came from a German reference (Hans Eberhard Fischer. 1963. Tetraploide Beta-Ruben Durch Acenophthen-Applikation. A Pflanzeng . 49: 91-95). If no one can find this reference, or having found it, cannot read German, I have a translation which was done by Sam Emsweller 11/14/63. I will not know the success of this method until later, but if it works, it is far less difficult than using colchicine. I don't think my 1986 efforts resulted in any tetraploids, but I am hopeful for 1987 because of the many plants and seedlings involved.
7.  Upon obtaining the above tetraploids (which are improved creams) and the nice yellow-flowered R. calendulaceum plants, I will then continue as outlined in 1963! - 24 years later.
8.  For your program you should have a picture of Pryor Yellow 75-305. I am not familiar with Pryor Red 75-315 that you mentioned in your letter, and unfortunately I do not have a picture of Pryor Yellow 75-305.
9.  I obtained a cutting of evergreen azalea 'Creamy Perfection', a plant of unknown origin from one of the gardens in Charleston, S.C. I was told it was light yellow, but I have never seen the flower. This clone might accidentally be the same as the one described by Nuccio in his 1986-87 catalog as Kurume azalea 'Mizuno-Yamabuki' - an old variety "lost" for many years. "Very, very creamy white almost yellow". I ordered this plant, but the stock was already sold.
10.  I am also attempting to develop a tetraploid R. keiskei from a very, very yellow seedling, with the idea that R. keiskei might work if R. calendulaceum fails. Anyone who has tried to double chromosome numbers will know how difficult it is. I am also still intrigued with the note put in my 1966 article in the ARS Quarterly Bulletin (Kehr, A.E. 1966. "Breeding for a Purpose", Vol. 20:3 pp. 130-141) that tetraploid roses are more intense yellow than diploid roses. Wouldn't it be nice to know if this is also true in azaleas and rhododendrons? I have some yellow rhododendron seedlings being treated with ACE (see (6).
I neglected to tell you the recessive white is 'Perle de Snynaerde'. I got this information by gleaning the writings of Dr. Heursel of Belgium - one of the speakers at the 1986 ARS Breeders Roundtable. We have no one in the U.S. that is doing work on genetics of flower colors comparable to what Dr. Heursel is doing. It is sad to say that probably neither the U.S.D.A. nor any state would support such work in the U.S. See pages 406-412 in Azaleas , by Fred Galle about Dr. Heursel's work on inheritance of flower color in azaleas.
Bob, this has been a rambling letter and I have told you all I know. Now anyone can breed a yellow-flowered evergreen azalea. As I said back in the dinner meeting at Ivan's Restaurant in the 1960's - "I think a yellow evergreen azalea can be created" - I still do.
Following is the article in the "Rhododendron", publication of the Pacific Rhododendron Society, to which Augie makes reference in the above letter.

A Potential Breeding Method To Produce A Yellow-Flowered Evergreen Azalea - August E. Kehr
A yellow-flowered evergreen azalea is the dream of many enthusiasts. As a result of the report of tetraploid forms occurring in the Satsuki evergreen azaleas, a method of transferring the genes controlling yellow flowers from deciduous to evergreen types is suggested. This present article describes the potential method and outlines the theoretical background. Although this method has worked in other plants and looks promising, it is not yet fully proven in azaleas. It presently therefore is hypothetical for azaleas. The idea is presented now only to stimulate efforts on the part of other interested persons to try their luck.

Diverse Species Crosses are Often Sterile or Abnormal
Frequently crosses between widely diverse species are sterile and useless for further breeding work. Essentially, the reasons are these: at the time of formulation of the germ cells, which later develop into pollen or ovules, the plant chromosomes in the germ cells do not divide but remain intact and pair with identical whole chromosome partners. After initial pairing, the whole chromosomes separate into two daughter cells. Thus the chromosome numbers in the germ cells are reduced to one half; in diploid azaleas the germ cells have only one complete set of chromosomes or a total of 13. At fertilization when 13 chromosomes from the pollen are combined with the 13 chromosomes of the ovules, the diploid, or 26 chromosome number, is restored. The essential requirement in the process of germ-cell formation is, therefore, that 13 chromosome partners capable of pairing must be present. If the 13 pairs are not able to pair, the process breaks down and few or no pollen grains or ovules are formed. In crosses between two diverse species, such as evergreen ( obtusum ) and deciduous ( flavum , also known as luteum ) azaleas, the F1 hybrids possess 13 flavum chromosomes and 13 obtusum chromosomes. In the process of germ-cell formation such unlike chromosomes would not be expected to pair, and, as a result, the occasional hybrid shows the effects of genetic unbalance, often including sterility. Most rhododendron fanciers are familiar with azaleodendrons, or F1 hybrids between rhododendrons and azaleas such as 'Glory of Littleworth' or 'Broughtonii ureum', which are sterile because of unbalance. This is also the reason that mules, so-called after the well-known breeding practice of a male horse with a female jenny, or jack with a mare. The resulting animal or plant cannot be used for further production of offspring. Plants which we call mules usually have no stamens. Such unbalances have been the barrier to many previous crosses between flavum and obtusum types in that the diploid F1 hybrids are albinos, virescents, dwarfs or other abnormals.

Diverse Species Crosses are Often Difficult
Anyone who has tried making crosses between obtusum and luteum types knows that such crosses are difficult. However, successful crosses are possible. For example, successful crosses have been made by the author as follows: 'Hinodegiri' x Mollis hybrid, Rhododendron mucronatum x Mollis hybrid, Glenn Dale 'Swansong' x Mollis hybrid, and 'Stewartstonian' x Mollis hybrid. Despite the few examples of successful hybrids, the hybridization of evergreen and deciduous types at the diploid level is far from successful from the standpoint of obtaining normal, viable hybrids.
In contrast, crosses between tetraploid types are often possible even when such crosses fail or are difficult at the diploid level. Many examples of such successful tetraploid crosses have been reported in plant science literature. However, tetraploid evergreen azalea crosses were not feasible until the discovery of tetraploid Satsuki forms. Although tetraploids of many plants have been successfully produced by artificial means, such as through the use of the drug colchicine, this method to date has not been successful with azaleas.

Tetraploid Azaleas Overcome Barriers to Inter-Specific Fertilization
In the luteum types there are two naturally occurring tetraploid forms, R. canadense and R. calendulaceum . It has been reported by F.P. Lee in The Azalea Book that Satsuki cultivars 'Banka', 'Wako', and 'Taihai' are tetraploids, and there may be others. Tetraploids usually have larger flowers and anthers, as well as petals heavier in substance. The cultivar 'Getsutoku', for example, looks like a tetraploid. In addition, some of the Hirado hybrids are tetraploids. A few of these new types are now available in very limited quantities at some commercial nurseries.
Crosses between R. calendulaceum and normal diploid evergreen Obtusums are extremely unsuccessful regardless of which species is used as a seed parent. Most crosses produce no seeds. In the infrequent crosses where seed is produced, it is seldom viable; or if it is viable, the resultant plants are not true hybrids.
The author has not yet had the opportunity of making the cross R. calendulaceum (seed parent) with diploid evergreen Obtusums as the pollen parents. From others it is believed that these crosses are seldom successful, and no hybrids are known. However, the cross R. calendulaceum (female) x tetraploid evergreen Obtusum (Satsuki cultivars 'Taihai' and 'Getsutoku') as pollen parents (pollen received from Plant Introduction Station, Glenn Dale, Maryland), was highly successful and large amounts of viable seed were produced. The seedlings when they first appeared were normal in appearance. By the time the first leaves were formed, a small number of the seedlings developed areas of albino or virescent tissue in some segments of the leaves, but growth appeared normal even in such affected plants. In subsequent growth, about three-fourths of the plants in the cross R. calendulaceum x 'Taihai' were retarded in their growth, while a lesser number of dwarfed plants were found in the cross R. calendulaceum x 'Getsutoku'. The non-dwarfed plants grew rapidly and developed as normally as comparable seedlings between R. calendulaceum and deciduous species. It is too early to know whether these crosses will produce true, normal hybrids. Cytological work is underway to determine chromosome numbers in the hybrids and to verify their probable tetraploid nature. It will be several years before the plants reach flowering age and the nature of the flowers is known. However, if such plants prove to be true hybrids, they would carry genes for yellow flower color and evergreenness and would be valuable for further breeding and backcrossing even in the event they themselves do not express both evergreenness and yellow color.
The germ cells produced by tetraploids normally carry two sets of chromosomes or diploid-chromosome compliments. If further research confirms that diploid pollen from the Obtusum parent combined with diploid ovules of the flavum parent, the tetraploid hybrid thus formed would be essentially a combination of two complete diploids in one single parent. Such special tetraploid plants are known as amphidiploids, diploid on both sides. As a rule, amphidiploids are highly fertile and normal.

Amphidiploids Offer Opportunities for New Types
Because they are fertile and produce viable pollen and ovules, amphidiploids are useful in the development of new and unusual types. In most cases they will cross back to one or both tetraploid parents. They will also frequently readily hybridize with diploid types. The important distinction between amphidiploids at the tetraploid level and similar hybrids at the diploid level is that amphidiploids are genetically balanced and seldom show the sterilities and abnormalities that limit the value of their diploid counterparts.
Amphidiploids have another important character in producing new forms. Because they have the essential chromosomes necessary for two distinct diploids, mutations and genetic unbalances that might cause nonviable germ cells in diploids are not likely to be detrimental in formation of germ cells in amphidiploids. Thus, a lethal character in one diploid set of an amphidiploid would be compensated for by the necessary "viability" factors in the other set of diploid chromosomes. In brief, amphidiploids can act as reservoirs for maintaining untold numbers of mutant and recombinant types, and hence they are a rich source of new variability. There is real hope for the development of yellow-flowered evergreen types in such a storehouse of genetic diversity and flexibility, especially in back-crosses with the tetraploid R. calendulaceum parent by selected amphidiploids that most nearly possess the evergreen habit. The potentialities of such crosses are great and they open new avenues for arriving at the goal of many azalea enthusiasts, that of producing a yellow-flowered evergreen azalea.
The second article which Augie refers to in his letter to me about acenaphthene follows:

An Experimental Method Of Inducing Polyploidy
For the adventuresome person who likes to experiment, another method to induce polyploidy may be tried, using a substance closely related to moth balls. So far, this method has never been used on rhododendron or azalea spp., but has been more effective than colchicine on plants such as sugar beets. The method is very simple, and so there is merit in its trial.
The chemical used is acenaphthene, which comes as white crystals very similar to moth crystals. Seedlings are grown in flats, covered with a sheet of glass. Just before the first true leaves appear, they are treated as follows:
A.  Dissolve the acenaphthene crystals in acetone or ether. Place the solution on the glass cover and allow the ether or acetone to evaporate, leaving a thin film of acenaphthene adhering to the glass.
B.  Replace the glass cover with the acenaphthene film side down toward the seedlings. The acenaphthene will vaporize (exactly like moth crystals) and the flat will become saturated with the vapors. Leave for three to five days, then replace the glass cover with clean glass. The temperature inside the flat is the critical factor in this method, and unfortunately no information is available on the optimum temperature for rhododendrons and azaleas. It is recommended that a good starting point is 75 degrees F. (24 degrees C.). A higher temperature increases the biochemical effect, while a lower temperature decreases it.
C.  Repeat the treatment in two or three weeks, except that the second treatment should last for only two to four days.
D.  If the rate of survival of seedlings is too low, decrease treatment times and reduce temperature.
This method has been successful because the vapor penetrates to the growing point so the growing apical point is affected. It also has the advantage of simplicity compared to the colchicine method. If you use this method, you are plowing new ground. Good luck!

My Comments On Yellow Evergreen Azalea Hybridizing
I suggest hybridizers consider the following:
A. For the deciduous azalea seed (or pollen parent) use the best of the Knaphill, Exbury or Ham clones in the intense yellow, yellow-orange or orange colors. Do not use Mollis clones. Reason: the Exbury and Ilam types, especially, contain upwards of eight to 15 species contributing to the genetic pool in the cultivars. R. occidentale was introduced by Anthony Waterer to the Knaphills. Some of the CW (collected wild) forms of R. occidentale have been preliminarily determined to be tetraploids. Azaleodendrons are easily produced with R. occidentale as the seed parent. I have seen such flowering plants using 'Margaret Dunn', 'Gold Mohur', 'Fabia', 'Purple Splendour', R. macrophyllum , 'Mrs. Donald Graham' and 'Loderi King George' as pollen parents.
Some of us in the Northwest Hybridizers Group, though not obtaining test chromosomal counts, believe many Exbury clones to be tetraploids. I suggest that the "yellowest" clones of the forms of R. occidentale collected by Britt Smith and Frank Mossman on the Oregon and California coastline be used for hybridization.
Is it not possible that some species of deciduous azaleas might contain a "barrier mechanism" to prevent easy cross pollination? This barrier could be mechanical or chemical. Rhododendron cultivars that produce little or no pollen at garden temperatures will yield fertile pollen when the budded plant is brought into a greenhouse where the daytime temperature reaches 75 to 95 degrees F. It has been noted also that non-receptive pistils become more receptive to pollen at such elevated temperatures.
Some of our Northwest Hybridizers Group members report successful fertilizations with unusual pollens by first emasculating an opening flower and waiting a day or two for the pistil to become moist and receptive to pollen. Then, using a clean razor blade, the style is cut cleanly just above the ovary. The moist liquids on the stigma are deftly transferred to the clean cut area of the remaining style. The desired pollen is placed on the stigma fluids, from whence it can easily reach the ovary. This reduces the chance of the stigma or style containing mechanical or chemical blockages to pollen tube growth.
B. I have seen the bloom on an R. kaempferi var. albiflorum sent from Mr. Hideo Suzuki of Japan. It is white, but with a creamy cast to the flower's center. It may be usable.
C. Augie's suggestion of using the lepidote species R. keiskei , should be most carefully considered. First, years ago during a visit to Hjalmer Larson's nursery in Tacoma, Washington, he made a comment I have never forgotten. Hjalmer told me that "when I get a new evergreen azalea cultivar cutting in the summer, I just green graft it on a lepidote species or cultivar such as R. davidsonianum , etc. They are compatible." I never tried it though, figuring he might be pulling my leg.
Second, around 1969-71, I first saw blooming plants of W.H. Hardijzer's azaleodendrons - 'Ria Hardijzer' ( R. racemosum (a lepidote) x 'Hinode Giri') and 'Hardijzer's Beauty' ( R. racemosum (a lepidote) x Kurume hybrid). Now it seemed Hjalmer was quite possibly correct.
I suggest attempting using the yellowest lepidote species with selected "yellowish" or tetraploid evergreen azalea cultivars. Species such as R. keiskei , R. valentinianum , R. hanceanum , R. xanthocodon , R. concatenans , R. lutescens , R. megeratum , R. luteiflorum , R. trichocladum , R. sulfureum and R. sargentianum to reproduce Mr. Hardijzer's type of azaleodendrons. Then backcross to the evergreen azalea cultivars. Or, use Mr. Hardijzer's plants with the yellow lepidote species.
D. In the 1950 issue of "The Rhododendron Yearbook" of the R.H.S., Dr. E.K. Janaki-Ammal and her associates reported the following chromosome counts. Most species have 13 pairs, 2N = 26 (diploid = 26, tetraploid=52, hexaploid=78).
R. concatenans 52
R. flavidum 78
R. ambiguum 52
R. xanthocodon 78
In other words, many yellow lepidote species in subsections Triflora, Maddenia, Cinnabarina, Lapponica and Glauca may already be natural tetraploids and can be used immediately with the tetraploid evergreen azaleas to produce our long sought after "yellow evergreen azalea" in the first crossings.
E. Last, the Robin Hill cultivar 'Olga Niblett' is one of the most yellow evergreen azalea clones I have seen, and it should be used. It contains R. kaempferi on both sides of its parentage and is probably quite rich in yellow flavonal pigments.

Bob Badger, a chemist by training and an avid hybridizer, operates Gnome Gardens and Landscaping with his wife, Marjorie. They are members of the Northwest Chapter, ASA and Seattle Chapter, ARS. Presented at the Ninth Annual Convention of the Azalea Society of America in Portland and Eugene, Oregon, April 27-29, 1987.