Theory and Practice
by Donald L. Hardgrove, Merrick, L. I., New York
Just for the sake of the record I would like to make a few corrections and some comments on articles appearing in several past issues of the Quarterly Bulletin of the American Rhododendron Society.
Mr. David Leach writing in the January 1953 Bulletin went into a great deal of detail in describing how I had, in the 1945 Yearbook of the American Rhododendron Society, given R. 'Meadowbrook' a three star rating, and that I considered it superior to such distinguished rhododendrons as R. 'Essex Scarlet', R. 'C. B. Van Nes', and R. 'Cynthia'. To further amplify his point he stated that I considered it equal to R. 'Souvenir of W. C. Slocock', R. 'Mother of Pearl', R. 'Dr. Stocker' and R. 'Mrs. A. T. de la Mare'. Inasmuch as Mr. Leach went to all the trouble to scout back to 1945 to prove a point, it is really too bad that he didn't delve a little deeper in the interest of accuracy. In the article to which he refers in the 1945 Yearbook I stated that my own ratings were given to a few varieties which had not previously been rated. R. 'Meadowbrook' however, had previously been rated in Dr. Clement Bower's book Rhododendrons and Azaleas published back in 1936. I merely used the rating in his book.
R. 'Meadowbrook' is a fine looking foliage plant and is indeed a very vigorous grower, but as far as being the equal of 'Souvenir of W. C. Slocock', 'Dr. Stocker' and 'Mother of Pearl', I would like to go on record here as not even considering it in their class. Had inquiries been made I would have been glad to supply my true opinion of R. 'Meadowbrook' before misinformation was presented to our members. In defense of Dr. Bower's rating it must be noted that nearly all of the iron-clad catawbiense hybrids are given merit stars for use in severe climates. 'R. Mrs. C. S. Sargent' is given 4 stars, 'R. Lady Armstrong' 3 stars, 'R. Catawbiense Album' 3 stars etc.
My garden contains one of the largest plants of R. 'Meadowbrook' in existence. Mr. Paul Vossberg who is propagating rhododendrons commercially, and is generally considered as one of the best propagators in the entire country, has full use of most of my plants to make cuttings. If, as has been claimed, Mr. Vossberg rates R. 'Meadowbrook' as a top notch variety, it is difficult to understand why he is not propagating it.
If, after all the confusing pros and cons, I may be permitted to dwell on breeding, I would like to give my own point of view based on actual experience. Mr. Leach, referring to the cross of R. 'Fabia' x (catawbiense x griffithianum), wrote as follows. "The very best that can be hoped is that the progeny will be in hardiness about half way between catawbiense and 'Fabia'." In several other instances Mr. Leach has insisted that some element of hardiness must be present on both sides of the cross if we are to expect any hardy progeny. Based on theory alone I might be inclined to agree with this idea, but based on actual experience I have found that it is not true. As an example let's take the cross of R. 'Atrosanguineum' x R. griersonianum.
Here we have a complicated hybrid on one side of the cross and on the other side a species which is completely tender. R. 'Atrosanguineum' is made up of a number of species probably including ponticum, catawbiense, arboreum, and thomsonii. I agree that only one of these combined individually in a primary cross with R. griersonianum will not produce any really hardy offspring, but R. 'Atrosanguineum' x R. griersonianum will, and did. True, the percentage of hardy seedlings was very low, but it was expected to be low.
Both Mr. Leach and Mr. Nearing have in past Bulletins made frequent reference to Mr. Nearing's cross of (decorum x griffithianum) x hardy red hybrids. The selected varieties resulting from this now famous cross are known as the Ridgewood hybrids. Here again we are dealing with a cross with hardiness on just one side. In the July 1952 Bulletin Mr. Leach referred to the Ridgewood hybrids as follows: "catawbiense x decorum is hardy under eastern conditions and this was merely a segregation of the semi-hardy decorum genes combined with the R. griffithianum flower size to produce in the intermediate progeny blossoms about the size of decorum and of a pink shade midway between the red of one parent and the white of the other. It was exactly what a geneticist would expect." Now why do I bring up this point? For several reasons. First, and perhaps most important of all, catawbiense x decorum is not hardy under eastern conditions, and this point alone refutes Mr. Leach's entire theory. Second, he is forgetting that we are not dealing with catawbiense alone on the male side of the cross. We are also dealing with R. arboreum, R. thomsonii and perhaps other species. When the split-up occurs in this cross, will any one of these species combined individually with either decorum or griffithianum result in progeny hardy under eastern conditions? I think not. Third, in my opinion, and I think in the opinion of a few others, decorum could hardly be classified here in the east as a semi-hardy rhododendron. I have tried many different collector's numbers of decorum and have yet to find a form which will reliably winter its buds, and bear in mind that my location is on, of the most favorable to be found in the East. R. catawbiense x R. decorum was definitely not bud hardy in Ridgewood, New Jersey where Mr. Nearing developed the Ridgewood hybrids. Now, inasmuch as Mr. Leach claims that it was this particular combination that produced the extreme hardiness of some of the Ridgewood hybrids, how does he explain the fact that R. 'Beatrice Pierce', R. H. 46-6 and several others of the Ridgewood hybrids are much hardier than catawbiense x decorum?
Again referring to the R. 'Fabia' x (catawbiense x griffithianum) cross Mr. Leach wrote, "To produce progeny which are hardy in the sense of being as cold resistant as the commercial catawba hybrids it would be necessary for hardiness to act in inheritance as a simple dominant in this formula. It has been abundantly demonstrated again and again by every hybridist that Rhododendrons are like other plant groups in that hardiness is a quantitative character controlled by many genes, some favorable alleles being dominant and some recessive." Mr. Leach is right in this respect, if we are interested in producing large quantities of hardy seedlings we incorporate hardiness from every direction. But we are not looking for thousands of hardy plants possessed of little beauty, we are looking for a small percentage of hardy plants possessed of much beauty. It is really hardly necessary to expound the theory that the more hardy elements we incorporate in a cross the larger percentage of hardy seedlings will result. Naturally in a cross such as R. 'Wellesleyanum' x (catawbiense x fortunei) we would expect nearly all hardy progeny, just as in a cross such as R. 'Tally Ho' x R. 'May Day' we would expect the progeny to be red. But the first cross would give us very little beauty and the latter very little hardiness. We try to balance the equation by putting in all the elements needed to supply what we want, and when the characters split-up and recombine as they will when two primary hybrids are crossed together, a small percentage of the seedlings which contain good hardiness will be possessed of the beauty of the tender varieties. It is in these individuals that new hardy varieties are founded.
Mr. Leach's article in the January 1953 Bulletin he has cited two crosses, R. catawbiense var. 'Glass' x (discolor x dichroanthum), and R. catawbiense var. 'Glass' x (fortunei x campylocarpum), with the following comments: "The semi-hardy R. fortunei genes will combine with the yellow color factor from campylocarpum to produce in the intermediate progeny hardy rhododendrons with pale yellow blossoms in size half way between those of catawbiense and of R. fortunei." Referring to the second cross his comments were. "Since we also know that R. catawbiense x R. discolor produces hardy progeny, such a cross as R. catawbiense album 'Glass' x (discolor x dichroanthum) will result in a segregation of the semi-hardy discolor genes with the orange color factor from R. dichroanthum to produce hardy progeny with good sized flowers of orange hue." All of which sums up to just about the most concentrated bit of wishful thinking I have ever encountered. What happens in these crosses to the tender genes of R. campylocarpum and R. dichroanthum, or are we to believe that these genes have suddenly become recessive and will not appear in any of the progeny? What happens to the heretofore recessive colors yellow and orange that we can count on them to appear in the seedlings, or have yellow and orange suddenly become dominant? (Oh how I wish they would!) What happens to the genes of the small sized flowers of dichroanthum and campylocarpum that we can count on flowers half way in size between catawbiense and fortunei, or in the case of the latter cross, "good sized flowers"? Have the small size genes become recessive and the large size genes dominant? Another very important point which Mr. Leach apparently did not consider is that we do not know very much about the behavior of R. catawbiense var. 'Glass' in hybrid form. Here we are dealing with a rhododendron quite new in the field of breeding. Can we assume that it will throw no purples or magentas? Mr. Leach has made such an assumption in his statement as to what his proposed crosses would produce. In fact his entire theory is built on the premise that this form of the species will produce no dirty colors when crossed with other rhododendrons. Frankly I do not know enough about this comparatively new form to make any claims either good or bad for it. This past season I had four of its hybrids bloom for the first time with the following results: R. catawbiense var. 'Glass' x R. fortunei produced flowers in color and size almost identical with those of R. catawbiense (type) x R. fortunei, there being a noticeable purplish tint. This plant was a real disappointment as I had planned on using it for breeding. R. catawbiense var. 'Glass' x ('Atrosanguineum' x griersonianum) looked a good deal like the old hybrid R. 'Caractacus', although here we can perhaps blame some of the purple on the influence of R. 'Atrosanguineum'. R. catawbiense var. 'Glass' x 'Fabia' produced small drooping flowers containing a large dose of magenta, and I sincerely believe it to be my biggest disappointment of the entire season. When the buds appeared on this plant last fall I started making a list of varieties with which to cross it. By the time the buds started to swell I had some fifteen varieties on the list, but when the genes of the insignificant, drooping, purple-magenta flowers manifested themselves the list was torn up. The one bright spot was the plant of discolor x catawbiense var. 'Glass' which produced pale pink flowers of fairly good size with no objectionable tint. Of course only one plant of each cross bloomed so no definite conclusion could be made. Unquestionably some of the other seedlings will be better, and in these we should look for the individuals to be used for further breeding. But it must be recognized that catawbiense var. 'Glass' may not contain the 100% white genes for which we all hoped.
Some of the seedlings of Mr. Leach's proposed crosses would undoubtedly be hardy. Some of them might carry yellow or salmon tints. But to say that the seedlings as a whole would carry all these good qualities is just not good genetics. In almost any cross involving three different species a segregation or split-up of characters occurs which will give us variable progeny. In the crosses mentioned by Mr. Leach the seedlings will not be uniform. Some will not be of sufficient hardiness to be useful in the eastern United States. Some will have the small flowers of dichroanthum or campylocarpum Some will be of poor color. Some will not have a sufficient number of flowers to make a complete truss. Some will have the flowers too drooping as often happens when breeding with the neriiflorum series, and some will not be floriferous enough to make good garden subjects. Naturally when making a cross we always hope that our percentage of the good qualities will be high and the poor qualities low. But we must never believe that a cross involving three different species will give us uniform results of all the good points we desire. To do so is purely an attempt to evade reality.
Another allegation made by Mr. Leach was that a cross between two different primary hybrids would result in progeny about midway between the two hybrids more or less the same as crossing two species together. To use his own words, "The formula presents no opportunity whatever for the appearance of any but characteristics roughly intermediate between the two parents and/or the four grandparents. In this respect it is similar in result to a cross between two species."
I have no idea where Mr. Leach could have obtained such information. Surely not from actual experience. Nothing could be further from the truth, as in actual practice the opposite occurs. In this type of cross the greatest split-up and recombination of characters results, with a wide variation in the progeny. This is clearly demonstrated in the nursery row with lined out seedlings of several of my own crosses. In the cross of (R. maximum x R. haematodes) x (R. orbiculare x griffithianum) a few of the seedlings look almost exactly like orbiculare having the typical dense habit and rounded leaves. Some have quite large leaves like R. griffithianum. A few look so much like R. maximum that if planted apart without labels one would almost surely say that they were the species itself. Some were killed outright last winter while others came through with no injury at all. In the cross of (R. catawbiense x R. haematodes) x (R. thomsonii x griffithianum) of which I have about 500 plants, similar variation was obtained with some plants appearing resembling thomsonii itself, some close to haematodes, and some near catawbiense. In the cross of (dichroanthum x venator) x (maximum x discolor) an extremely wide variation was obtained, and incidentally several very striking plants. The seedlings of (R. maximum x R. haematodes) x (R. repens x R. griersonianum) also show evidence of being widely split.
Does this sound like the appearance of nothing but intermediate characteristics? Does this sound like a similar result to a cross between two species?
If further proof is required I would like to quote a hybridist with many years of experience and whose reputation is beyond question. Mr. F. C. Puddle, in the 1998 Year Book of the Royal Horticultural Society, wrote as follows: "Probably the most varied result which I have obtained from any cross was in R. 'Wilfred' (williamsianum x 'Phidias'). The colors of the seedlings included cream, yellow, orange, pink, and red in fact, scarcely two plants were alike. If we search for the reason by dissecting the pedigree we find that it includes the following species, R. griffithianum, campylocarpum, dichroanthum, griersonianum, neriiflorum and williamsianum thus all the colors in the elepidote series are combined in the one hybrid. Small wonder that the family is so varied."
My own experience with crosses involving three or four species has been not only the wide variation in form and color of flowers as Mr. Puddle has described, but also in habit of growth, hardiness, foliage, size of flower, form of truss, in fact in almost all characters.
A very interesting bit appeared in the 1949 Rhododendron Year Book of the Royal Horticultural Society. Particularly interesting to me as it is almost identical in thought to the ideas I expressed in my article in the 1946 Year Book of the American Rhododendron Society. The statement to which I refer appears on page 99 as follows: "You said it was desirable to cross two species. Now, Mr. Rothschild I know was very strongly of that opinion, and repeated it again and again. It is desirable from this standpoint: if you want a lot of fine progeny and very few poor ones, then it is desirable. But when you cross two natural species you get a progeny which is rather uniform, if those species are fixed and definite species. If they are species in a state of flux, such as those in the neriiflorum series, you are likely to get more variation, but on the whole you get rather uniform results. But on the suggestion of some British writers I made a study of the possibilities of secondary crossing, and it is my firm opinion that the way to get a supremely fine result is to cross, not two species, but having crossed two species such as R. campylocarpum x R. fortunei, to cross two other species such as R. griersonianum x R. eriogynum, and then to cross those two primary crosses together: and when you do that you get the Mendelian split-up in which all the characters are recombined. I said all-I should say many of the characters, because some of the characters apparently will not distribute themselves; but you get a recombination of characters, with the result that some of your seedlings are utterly worthless, many of them are far below average and they are in many cases totally different in all respect, but you have the possibility of a supremely beautiful result in some cases. You must expect to throw away 95%. at least of the progeny, but of the remaining 5% you are likely to get something very extraordinary, and something very different from anything you have had before."
One other point which has been freely discussed by both Mr. Nearing and Mr. Leach is that of the so-called "racial norm." In my article, Creating New Rhododendron Hybrids, published in the 1946 Year Book of the American Rhododendron Society, I stated that the crossing together of too complex hybrids would result in a racial norm with not much chance of anything good. For this utterance I was accused of being merely an echo of Mr. Nearing's statement. A very strange echo that originated 6 years before the actual sound, as my article was written in 1946 and Mr. Nearing's in 1952. However, my statement referred to the crossing together of the old R. catawbiense hybrids and referring to the results of such crossing as a racial norm is sheer flattery. A more accurate description would be "junk." Mr. Leach's citing of R. 'Meadowbrook' as an example of what can be expected from complicated crossing is well taken as R. 'Meadowbrook' is not anywhere near as good as R. 'Mrs. C. S. Sargent' one of its parents. And remember it was selected as the lest of countless seedlings! I wonder how much money was spent and how much time consumed in raising all the seedlings necessary to produce this rhododendron which is not as good as its parent.
I hope we can finally put the complicated breeding question to rest once and for all. To achieve that end I would like to offer a few opinions other than Mr. Nearing's or my own.
Thirty years ago Mr. J. C. Millais wrote as follows: "Those who have had any experience of raising hybrid rhododendrons know too well from bitter experience that the percentage of successes is very small compared to the failures. Many growers abandon the work owing to the hopelessness of producing anything better than what has already been done. One famous hybridizer said that if he obtained two good plants in 1000 seedlings he was well pleased; and quite recently Mr. C. B. Van Nes has told me that he raised 250,000 seedlings from second and third cross R. auklandii hybrids mated with certain scarlet hybrids, and that he retained exactly 200 plants, the rest being destroyed. So small a percentage of good things might seem to he an excessive case of failure but in reality it was not so. In fact, it was a great success; for the results were far beyond Mr. Van Nes's highest hopes, since amongst the lot retained were some of the most beautiful hybrids ever seen. It was, in fact, only what might be expected from breeding with triple or quadruple mongrels.
It is now a proved fact that when two good species, somewhat nearly related, are mated, the results show a very large percentage of success, just as the further we pursue the interbreeding of crosses, the greater are the number of failures, due to the loss both of individual characters as well as color and form."
Here we have a case where in final analysis perhaps only 20 out of a quarter of a million seedlings ever reached the market as popular named hybrids. This figures out to be one seedling in every 12,500. In the triple and quadruple mongrels used, was the blood of that superb rhododendron, R. auklandii, now known as griffithianum. If only one out of every 12,500 was good enough to become popular with auklandii somewhere in the background, can you imagine what the percentage would be with the auklandii blood taken out? And remember, the R. catawbiense hybrids have nothing like the magnificent auklandii in their pedigree.
For a bit more proof I would like to quote from a letter received from Mr. Joseph B. Gable back in 1940, "There are some things like R. catawbiense hybrid seedlings crossed inter se that I would now be willing to junk in the seed envelope. I have yet to see one worth while among the thousands that have flowered here."
From Mr. Paul Vossberg, "I believe I have raised well over 50,000 seedlings of the catawbiense hybrids and have yet to see one good enough to displace any of their parents."
The debate carried on in past issues of the Bulletin by Mr. Nearing and Mr. Leach has, however, served a wonderful purpose. It has brought to the attention of the average rhododendron enthusiast some idea of the problems encountered in breeding, and perhaps some idea of the amount of work and planning needed to produce new hardy hybrids. I have no intention of entering into a breeding debate with Mr. Leach, Mr. Nearing, or anyone else. I believe that a debate on the subject of theory alone is inclusive and inadequate to those who are thirsting for facts. It is my intention to report these facts. learned from actual results, as quickly as they occur or are reported to me, in the hope that they may be utilized to improve the results of rhododendron breeders.
Mr. Leach I hold in the highest esteem for his sincere effort to inject new ideas in the development of hybrids. Mr. Nearing I am proud to have as a personal friend. I feel sure that some of the theories proposed by Mr. Leach will be helpful to all of us, but my primary purpose in this presentation is to make sure that theories remain theories and are not presented as, or confused with actual results.