JARS v36n1 - The Quest
David G. Leach, North Madison, Ohio
Editor Ed Egan has asked me to write an article about my hybridizing goals and the efforts that have been made to realize them. It is not an easy assignment.
The aims have shifted through the years. Discouragements have stifled some. New needs have arisen in landscape architecture. Many of the older practices have become obsolete and they are not likely to have a cyclical revival. In working with a genus in which progress is so slow, it does seem that rhododendron breeders should try to have a vision of tomorrow's lifestyle, and I like to think I can foresee a future demand for some characteristics in the future which are not now sought. So the concepts have evolved in fluid progression with the passage of the years; all that can be done is to reverse the winding reel and freeze a frame of the action for a moment as it appeared at the time.
Looking down the alphabetical list of hybrids that have been named, some illegitimately by inadvertence, I see at the top, 'Anna H. Hall'. It was solely the product of a single objective: to produce something close to R. yakushimanum which would be hardier, and vigorous enough to be a practical commercial rhododendron. I wanted to retain most of the charms of the famous species from Yaku Shima which is not, alas, a profitable plant for nurserymen to produce. 'Spring Frolic' came from the same lot, as did 'Great Lakes' and 'Pink Frosting', all the offspring of a cross with R. catawbiense var. album , and all blooming at different seasons.
The quality of these first generation R. yakushimanum hybrids, produced long before this species became generally available for breeding, was a fortunate fluke. Many other crosses produced progeny inferior to either parent. It is only now, in the fourth and fifth generations, that I am starting to approach the goals I had hoped to obtain in the first.
There followed next a group of hybrids, of which too many were named perhaps because they represented a success after so many failures, from attempts to capture the attractions of R. williamsianum in forms which would be adaptable to the eastern United States. All of the early hybrids were unable to endure the aridity of eastern summers. Finally, Dietrich Hobbie gave me pollen of 'Adriaan Koster' x R. williamsianum , and that, crossed with R. catawbiense var. album , produced a fine batch of semi-dwarf white to pink flowered, sturdy, heat resistant hybrids, with one out of 400 a frail, pale yellow. At the time they were named, 'Finlandia' (syn. 'Alaska'), 'Robin Leach', 'Flair' and 'Applause' seemed sufficiently different to merit christening. 'Finlandia' is the most distinctive of the lot, with small, sharply convex leaves which produce a foliage texture new to rhododendrons in the East. It is extravagant in its production of glistening white, fully filled flower trusses. Landscape architects are attracted to its unusually trim, tiered but dense foliage, the leaves overlapping like small inverted green saucers. All of these bloomed after -28° F. following the winter of 1962-63 in the mountains of Brookvìlle, Pennsylvania.
R. 'Robin Leach'
photo by David Leach
Having seen the dazzling dwarf scarlets in England, there was then an instantaneous compulsion to produce their equivalents in hardy form for the arctic climate of the western Pennsylvania uplands. It turned out to be a long and often discouraging road. The first generation R. forrestii hybrids were all bud-tender, and often plant-tender as well. Once again, Dietrich Hobbie's work provided the key that opened the door. His 'Gertrud Schäle', a hybrid between R. forrestii and 'Prometheus', crossed with 'America', my own 'Fanfare' and others, produced a large population of semi-dwarf scarlets. Through the years they were slowly winnowed down to 'Small Wonder', 'Sumatra', 'Flamenco', 'Rangoon' and 'Singapore', five out of perhaps 2,000 seedlings but still too many. 'Sumatra' is virtually faultless as an evergreen, with handsome, dense foliage on a plant three feet tall and five feet wide after 24 years. It blooms in tumbling tiers of scarlet bells, not so much in trusses as in a cascade of incandescent pendants. 'Small Wonder' is the hardiest of the lot, also a new standard of color purity for the East, but a little more open cushion eight feet across and four feet high after 24 years. It should be a valuable parent, with its dwarfness, electric scarlet flowers and hardiness. However, at least one nurseryman has told me that it "makes up" too slowly to be a profitable rhododendron commercially. I tend to think it may bring a price premium on the market that will justify its slower growth, from those who resonate to red. It is distinctively different from anything now commercially available in cold climates. Then, of course, there were the lovely clear pinks of the West Coast and England, so appealing in contrast to 'Roseum Elegans' and our other blue-flawed pinks. They were easy, right from the first, to produce in "ironclad" form. The tender 'Pilgrim', 'Mrs. Furnival' and others were open handed contributors of their colors untainted by mauve. 'Bravo' was one of the beneficiaries, a huge, blowsy, Rubenesque sort of rhododendron with buxom trusses of not very distinguished carriage. There have been many more: 'Persia' is my favorite, a very pale pink of great delicacy and distinction which is so floriferous that a field line-up of six-inch transplanted rooted cuttings will each be crowned with a pink puff at the end of May. The hardiness in all of these clear pinks came from the most effective and reliable source of all, the white Catawba rhododendron ( R. catawbiense var. album ).
photo by David Leach
R. 'Small Wonder'
photo by David Leach
Even more appealing were the so-called "art shades" by which was meant the blends of pink with yellow and orange. None existed for really cold climates, so the challenge was obvious and irresistible. I sought as parents for color contribution hybrids which included in their ancestry R. neriflorum , R. dichroanthum , and at least one species in the Fortunea Subsection (Fortunei Subseries), the latter to produce a flower more flared, and to provide a greater degree of adaptability to a variety of climates and growing conditions. The theorem proved to be reasonably valid. An array of yellow-pinks and orange-pinks appeared through the years which were a startling novelty in a 20-below-zero climate. Several were released that later demonstrated their unsuitability to commercial production; they will always be hobbyists' hybrids. But 'Bali', 'Bangkok', 'Nuance', 'Peach Parfait' and one or two others yet to be named propagate readily, grow vigorously and seem to have no problems that would disqualify them from large scale field cultivation.
photo by David Leach
It is at least generally true in breeding rhododendrons that like does tend to beget like. A parent that is hard to root, for example, is likely, more often than not, to produce progeny that are hard to root. This seems to apply more regularly with the intangible characteristics. A parent that blooms partially in the fall is apt to yield offspring with the same defect. Precocious flowering, or the lack of it, especially, is prone to appear in successive generations. A conspicuous exception to the pattern is a deficiency of vigor; hybridity, or an extension of it, often remedies a weak constitution evident in a parent.
Whites of large stature have not been, for me, such an enticing goal. The old hybrid, 'Catawbiense Album', produced by Anthony Waterer in England more than a century ago, is a good rhododendron, widely distributed and extremely hardy. My efforts have been largely concentrated on creating cultivars which bloom both earlier and later. An extremely early flowering clone, 'Last Hurrah' (syn. 'Athens') is a semi-dwarf which was produced by a cross of R. aureum ( R. chrysanthum ) with 'Belle Heller'. It performed well for years, bearing large trusses of clean white flowers at a bleak season. No sooner was it named and distributed than it began to bloom partially in the fall. Now it blooms so heavily in September that visiting nurserymen comment on the ease of promoting it as an autumn flowering rhododendron. Apparently, there is a reduction of the pre-flowering dormancy requirement which accompanies maturity in this clone. The fall bloom is so heavy that there is scarcely any spring flowering in some climates.
'Belle Heller' crossed with 'Catalgla' produced hybrids predictably superior to 'Catawbiense Album'. They propagate much more easily, grow more vigorously, have better foliage and larger flower trusses more freely produced. Professional growers say they are a good deal more profitable to produce. 'Lodestar' is the more vigorous; 'Swansdown' has the edge for growth habit and foliage density.
'Summer Snow' was the result of an effort to extend the season on the late side for white hybrids, but its usefulness is very limited. It is a tree rhododendron, which will eventually grow in an arboreal manner to at least 30 feet. In late June it produces trusses of huge, opulent flowers which are both fragrant and of heavy substance. 'Summer Snow' needs a sylvan setting to look comfortable in the landscape. The leaves are very large and it is extremely hard to propagate. The parentage was a superior form of R. maximum crossed with a second generation hybrid between R. ungernii and R. auriculatum .
More effort has been made to produce yellow hybrids, with less results, than any other flower color. Thousands of seedlings have been grown from scores of different crosses. They have been yellow, but most of a pallid shade far from the 'Hotei' or 'Crest' that I sought to produce in "ironclad"-hardy form. Too often the foliage has been dull olive in color, or sparse, or subject to spotting in warm climates, or the buds have not been cold hardy. 'Prelude' has been much the best of the mild climate parents to produce hardiness accompanied by a modest degree of yellowness.
Looking back now, I think the work was done too much on a trial-and-error basis, and too little on the probabilities suggested by increases in our knowledge of rhododendron pigments. We have known since the mid-sixties that a flavonol called gossypetin yields yellow flowers, and this soluble pigment usually produces in yellow flowered wild species a luminous, somewhat translucent meline effect. R. wardii is an example. There are other, insoluble pigments, the carotenoids, in the corolla tissue, the presence of which can generally be identified by the opaque flower appearance, without luster or gloss. The carotenoids are responsible for the matte yellow flowers of R. lacteum , for example, and the orange of R. dichroanthum . In retrospect, it seems obvious that a sensible approach would have been to cross yellow flowered rhododendrons which would reinforce the color saturation of the progeny by deriving their yellowness from the two different pigment sources. Probably such examples as R. wardii or 'Crest' are the yellowest that the flavonols alone can produce. Ideally, reinforcement by carotenoids to the extent that the luminosity is not lost would yield a combination of transmitted yellow light via the flavonols and reflected yellow light from the carotenoids. There would be a tradeoff of some brightness and vibrancy for greater color saturation.
Such an approach was suggested by Santamour and Pryor in an article published in the October, 1973 issue of the Quarterly Bulletin . Although the disappointing progeny of 'Hotei' crossed with 'Crest' do not bear it out, in other practice it has been extraordinarily effective for Robert Furman, whose hybrids are, however, designed for the mild climate of a waterside garden on Cape Cod. There may be a much tougher problem for breeders in colder climates because gossypetin and the anthocyanins which normally produce the pink-red-violet-blue range, are so similar in chemical structure that there is a tug of war between the two for the limited amount of materials available within the plants for the synthesis of the yellow gossypetin. In the endless quest to produce hardier hybrids, breeders in cold climates may be adding, through R. catawbiense , its color forms and hybrids, extra anthocyanins which diminish the production of the yellow pigment. The information available suggests that a clear white R. maximum , whether as a species or in a hybrid, could be the best source of hardiness.
The competition between the red to blue anthocyanin and the yellow gossypetin seems not to have been studied in rhododendrons. At the end of a fine article on rhododendron pigments in Contributions Toward A Classification of Rhododendron , published by the New York Botanical Garden, Dr. Jeffrey B. Harborne gives a very good bibliography, listing his own and the publications of others on this subject.
Even with the pragmatic approach in breeding, a slow but steady progress in yellow flowered hybrids over 35 years has resulted in the naming of nine cultivars in that color class, three of which are ivory. 'Good Hope', a cross of R. catawbiense 'Rubrum' with R. wardii provided, with its siblings, valuable information on the inheritance of yellow gossypetin, and it is probably the best commercial hybrid in the total of its assets, though it is not the yellowest. In hot climates, it is disfigured by a brown spotting of the leaves. 'Hong Kong' ('La Bar's White' x 'Crest') is, to my mind, a rhododendron of style and distinction, but it requires full field exposure to be shapely. The large flowers are not carried in formal, upright trusses. 'Calcutta' is very yellow indeed, but the flowers are tubular and it is a sprawling grower; it has not been released. There has been virtually no competition from yellow hybrids already in the trade in sub-zero climates. The tender 'Goldsworth Yellow', the closest approach, must have been given its name by the most elaborate of courtesies.
R. 'Good Hope'
photo by David Leach
Wolfgang Spethmann, in his paper printed in Contributions Toward A Classification of Rhododendron , remarks that "orange color is found only if chromoplasts are found with anthocyanins in the epidermis." This corresponds with my experience exactly. [ R. maximum x R. catawbiense ) x ( R. dichroanthum x ( R. discolor x R. campylocarpum )] produced for me a striking orange which I called 'Poppinjay' for convenience in further hybridization. It has dull, undulant leaves, unfortunately, and it is not too attractive when it is not in bloom. Several orange flowered hybrids have been moved from the field to landscaped sites for further observation, and I feel now that it would not be hard to produce an array of orange flowered hybrids in a variety of statures and sequences of bloom.
One orange which will certainly be released resulted from the cross of 'Russell Harmon' (a natural hybrid of R. maximum and R. catawbiense 0 with 'Goldsworth Orange', and that hybrid in turn was crossed with 'America' x 'Gertrud Schäle'. This produced an emphatically orange hybrid with large, full, conical trusses which bloomed fully after the winter of 1976-77, the coldest in the 103 years of the Cleveland Weather Bureau. It roots readily and so far exhibits no problems of any kind. It is a lusty grower, however. I tend to think that any rhododendron hybrid produced today should mature at about five feet if it is to be useful and popular long into the future.
Finally, the late flowering elepidote hybrids: my introducers tell me with considerable emphasis that there is no market for rhododendrons which bloom after Memorial Day, when traffic in the garden centers abruptly declines. This may be true, but I think the fault is with the marketing, not with the blooming period.
Americans have become a race of outdoor dwellers in the summertime, broiling steaks on their patios and plotting for next year's swimming pools or perhaps tennis courts. It is hard for me to believe that, someday soon, they are not going to want some showy color from woody plants, and if that is true, what else but rhododendrons? With that conviction in mind, I have spent a lot of time breeding rhododendrons and azaleas which will bloom right on from springtime to the end of July. No commercial grower has given me the slightest encouragement; rather, there have been puzzled expressions suggesting doubts as to the judgment of any breeder so misled.
Perhaps the historic aversion to late flowering rhododendrons has not been solidly based. It seems to me that the real problem is floriferousness. A rhododendron or azalea with terminal flower buds on most of its branches is going to make a brave show regardless of its blooming season. In many cases of midsummer blooming hybrids, lateral growths do not start until after the buds have flowered. Even a well grown August blooming R. prunifolium is a pillar of fire in its season. And so it is with rhododendrons. R. maximum 'Mount Mitchell' crossed with ['Mars' x ('Mars' x R. catawbiense 'Rubrum')] produced a stunning array of July blooming rhododendrons of a scarlet more brilliant than any midseason hybrid. 'Summer Snow' x ( R. maximum x R. catawbiense ) yielded a batch of huge pearl pink, July blooming hybrids which, I am assured, will have no commercial future, ever, because there is so little patronage of garden centers after the spring planting fever subsides at the end of May. I don't believe it; their time will come, I think.
The same conviction prompted the creation of the Madison Group of late June blooming deciduous azaleas, mostly hybrids of native species, and of five named clones of mid-July blooming R. prunifolium hybrids. All of these make fine displays here and are so admired by visitors that many inquire where they can be bought.
Scaly leaved hybrids have not prodded much activity from me until the last six or eight years. There was an obvious need for a white flowered P.J.M. and it was no trick to produce one after plants had flowered of a white R. dauricum grown from seeds sent from Japan by Dr. Rokujo in 1962. Local nurserymen who saw the seedlings in the field urged the selection of the fastest, most vigorous grower because such a clone will "make up" faster and be more profitable to grow. I called such a plant 'Yukon' and released it, but I now believe the basis of the selection was a mistake. Propagated plants of a somewhat slower, much denser grower are, at four feet, a good deal more handsome. There may be a lesson here for other breeders, as there has been for me. Another factor is that virtually all hybrids derived from R. carolinianum with R. dauricum crosses seem to become more lanky, open growers in the South than they are in the North.
photo by David Leach
For the most part, my intermittent efforts to produce distinctive, hardy, scaly-leaved hybrids were failures until I obtained a semi-hardy form of R. keiskei from Mount Kuromi, on the same island whose giant volcanic grunt gave us R. yakushimanum . This was the final acquisition in a long line of R. keiskei accessions from various parts of Japan. It is low growing and, unlike its predecessors, it is compatible in crosses with many other rhododendrons. Mated with R. dauricum var. album , R. carolinianum and its variety, album , 'Wyanokie', 'Laetevirens' (syn. 'Wilsoni') and other such hardy parents, it has produced a gratifying progression of early flowering, pale yellow to salmon pink hybrids, many with winter foliage of interesting color. These are vigorous enough to be practical commercial rhododendrons; they languish, however, if the drainage at the planting site is sluggish. Some of them are fertile, an unexpected and important bonus.
R. chryseum x R. carolinianum produced an attractive, floriferous hybrid with orange buds opening to yellow flowers, which has already been named.
Gordon Emerson crossed R. keleticum x R. carolinianum , and I grew on a few seedlings, one of which is a neat, stylish little dwarf which produces a helmet of crisp, lavender-pink flowers every spring. This hybrid has the misfortune to bloom at the wrong season; the same floral effect, without the style, can be obtained at a fraction of the cost from an evergreen azalea. Breeders who live in green, as opposed to ivory, towers are prudent to weigh the competition at the estimated blooming time of their projected hybrids, especially those which will be more intimate in their floral effect than the massive displays of the conventional elepidote rhododendrons.
I remember very well that Joe Gable, Guy Nearing and others who helped me so much, were asked to write accounts of their rhododendron breeding after they had accumulated about the same amount of experience that I have. It is none too exhilarating now to join the ranks of my betters; I hope, however, that some readers will pick up a signpost or two to aid them on the tortuous path toward effective hybridizing.