Logo for the Journal American Rhododendron Society

Journal American Rhododendron Society

Current Editor:
Dr. Glen Jamieson ars.editor@gmail.com


Volume 33, Number 2
Spring 1979

DLA Ejournal Home | QBARS Home | Table of Contents for this issue | Search JARS and other ejournals

Plant Breeding - Understand Your Problems
Ernest H. Yelton, M.D., Rutherfordton, N.C.
Reprinted from 1976 Hybridizers Roundtable ARS Convention, Philadelphia, Penn.

        I am going to speak as a representative for probably the 99% of breeders who have predictably poor results.
        After making sort of a survey in all of my efforts in crossing rhododendrons over the years I went over the lists to see what I turned up with. I was rather shocked to find that I had very little to show for it. Feeling that probably I had not benefited from other people's mistakes- perhaps I could allow you to benefit from mine. So I got out a book that I had delayed reading for some time on plant genetics published in England and authored by William Watkin which I would highly recommend to you as a source of basic genetics. I am not speaking of the very advanced details of it.
        The principles of plant breeding are the principles of evolution. What we are seeking to do is create variability and variability occurs very slowly in nature. It has taken millenniums to produce the various species of plants that we have. New species are being formed right now somewhere in the Himalaya Mountains. So what we are trying to do is hasten time. That is literally a stupid thing to expect much of. We are dealing with a search for variability among 13 genes in rhododendrons as that is the basic genetic number of rhododendrons. Each gene has a potentially large number of forms of alleles or variations of the gene. In each gene there are quite a number of chromomeres which are the very finely subdivided portions of the gene. Each one of which are the very finely subdivided portions of the gene. Each one of these can be variable. So odd-infinitem this goes on. The potentiality for exponential explosion of these is so staggering that, as Gus Mehlquist told me, if you have 10 variables, than it gets to the point of infinity.
        So we're dealing with an enormous problem. The chances of things happening that you want are very slim indeed. You must accept what comes along and not be discouraged.
        If your concept of genetics is the same as mine was after I finished medical school, you always think of the old pea story. The Mendelian idea that you have one gene for one color and one gene for another color, height, stature, etc. It is a one for one basis, that there is all or none on that matter. That is a very simplified concept of genes. For example in flower color alone, it is estimated that in most plants, at least 4 or 5 genes are concerned with it, not just a single one. Further, each one of these genes is associated with much more than just the production of color. It is associated with stature, disease resistance, the plant habit and type of stamina that it has to survive. This is tied closely within the same gene to color. So you are not dealing with the isolated thing of color. The things unseen are often much more important than the things we recognize as genetic qualities of a plant. Just because we see a certain color flower, means that this is not directly tied to one gene; it is tied to many genes and many other things are tacked on to it.
        I have a little pet theory, for example, that the color yellow is associated with very strong pollen production; practically all yellow plants are great producers of pollen. On the other hand, I think red is associated with very poor pollen production. And so you see there is more interlocking and tie ups than we thought before.
        The reason for this is that each gene effects one or more enzymatic processes in a plant. The DNA-RNA concept of inheritance has come into play now and we are just beginning to understand what the plants knew all along that the biochemical processes of the plant are inherited and this is the basis for the evolution process.
        No single gene has an effect confined only to the character by which the action is most easily recognized. The character that we are looking at is not the only thing that that gene is affecting.
        The bad part of trying to work with plants is that the very traits you want to obtain are usually recessive. They are the variable traits, the things that have not become dominant in a plant because the dominance is already present. What you want to do is create a change and the gene that you usually want is recessive in most instances. So the recessive gene is the one that you'll have to do more work with. You must replace the dominant one with the recessive one. And that requires doubling or pairing of the recessive gene before it will become evident in the plant. Dominance of an unwanted gene can negate the good sought from a recessive gene. Mutated genes usually are recessive. The very ones that we want are the ones we'll have to dig out the hard way.
        A gene is modified in its effect by other genes through the cytoplasmic environment. This accounts for the importance of the female parent in a cross. The female is contributing the juice to keep this whole thing going and thereby, as is in my household, the female rules the show. So you assume that this is going to be the rule of life and live with it and where you can, you must try to allow the female to be the one with the desirable trait that you are trying to bring out.
        Gene actions change with environmental changes. I think we are just beginning to appreciate that we must go back to the Himalayas or wherever these things come from and seek out the one bush that is a little higher, drier or tougher looking, the one that was overlooked because the ones in the valley had the better flower than the ones on the mountain. They were also more tender. Environment effects the measurable things such as rate of development, size and number of organs. It can also effect the hardiness of a plant greatly. We must consider in selecting plant material where it came from.
        Each gene has a mirror image. In the diploid, for example, they are matched exactly in a certain alignment. In order for this chromosome to be effective, it has to be directly aligned. If it isn't, you run into all sorts of problems such as sterility and various lethal traits; the seeds might germinate, then die out after a year or so or might have no chlorophyll or none of the internal processes on the plant work properly. This explains, I think, the great problem we have in trying to convert the hardy lepidotes as parents for all these fancy R. maddenii, R. edgeworthii etc. species because we are trying to put polyploids on diploids. In order to survive this miss-match, we must create polyploids as Dr. Kehr has done. I would have saved myself much time, for example: in 2 years time I did 52 crosses on R. carolinianum, R. chapmanii and R. keiskei and other lepidotes in which if I had had tetraploid forms, perhaps I would have succeeded. I ended up with nothing. Polyploid species are able to withstand mal-positions of chromosomes, deletions of genes and duplication or translocation of genes. You get a better chance at it although it doesn't always work.
        The cytoplasmic effect on the female plant is important. For example in camellias, we all know that if you want a variegated flower, all you have to do is infect the plant with the virus that causes variegation. Practically every camellia has variegated forms; they are very beautiful, but they are the effects of the virus on the gene rather than being a genetic factor.
        This happens in rhododendrons too, I am sure, as it happens in practically all plants. I've never seen a variegated rhododendron but I am sure there are other effects, other than variegation, that are related to some virus effect.
        Incompatibility is the result of the inability of pollen tubes to penetrate the full length of the style to achieve fertilization. I suspect it is the inability of the pollen tubes to have enzymatic action. They don't have the will to live, you might say. Incompatibility explains why yakushimanum FCC selfed does not produce seed. I've tried that. It was a mistake - I should have read the book before I tried it.
        The yak is self incompatible. Now this doesn't mean that you can't take a yak that you got from the seed exchange and cross it with the FCC form. You'll get seed and lots of it. You can get seed from crossing FCC x the Exbury form. It just means this: selfing is what is producing the incompatibility. Just because you like a plant and you stick pollen on it the chances are good that you won't get seed at least from its own pollen. There are plants that are self compatible but there are lots that are self incompatible. The native American azaleas are self incompatible. I found this when I tried to self desirable forms of calendulaceum and others and they are just not self-able.
        What happens? Nature understands Mosaic Law - (Moses found out Nature's Law) you should not marry close relatives. This is what these rhododendrons are trying to tell us. If we are aware of this, we can be successful a lot of times when we've been failing. Over 3,000 plant species are self incompatible.
        I have a little chart, taken from the book showing a little secret that may help. If the female parent is self incompatible and you cross it with a self compatible male or an incompatible male the cross will usually fail-not always, but a good percentage of the time. If the female is self compatible, and you cross it with a male parent which is self incompatible the cross will succeed. So you want to have female parents that are self compatible. What does that mean? It means that if you are going to work with yak as a species, you ought to try to find one that is self compatible if you can. Because, then you can succeed much more quickly. I am talking of dealing within the species. I am not talking of going outside the species.
        Inbreeding produces homozygotes. A homozygote is practically a mirror image of its sister seedlings. Most time we want heterozygous plants because these allow the variability that we are seeking. If you have no variability, you are stuck. You just keep going to a dead end. Further, nature doesn't like the business of homozygotes too well. In-breeding will lead to suppression of plant vigor and it causes what is called in-breeding depression. And this happens within 2 or 3 generations of doing this. For example, a popular plant now is "Trudy Webster". If everyone like me (I read what's going on and copy someone else's work) goes out and gets 'Trudy Webster' and decides to try it "one more time". Well you keep that up you are going to run into some other bad things. The chances for bad things are far better than the chances for good things. You usually end up with poor yield, some changes in the height, width, winter hardiness-things that are pretty important if you want the plants to survive. These are called the metric characters of a plant or the measurable characters. In-breeding doesn't usually effect the meristoic characteristics such as the number of flower parts, the node length, color etc., and interestingly enough the way we arrive at what is a species is the meristic characteristics of the plant rather than the metric. The meristic characteristics are what you look for to identify a species. One wouldn't solely measure the height of a plant and determine from that in what species the plant belongs.
        We have had some comment in the past about breeding for disease resistance. I'm very concerned about this because the fatality rate of plants with root rot in my area is just phenomenal. Our temperatures in the summer and the humidity are real aids to this problem. A big problem in breeding for disease resistance lies in the fact that you are dealing with another viable substance and actually to some degree fungus injections also are plants. So when you are working one plant against another to vary the resistance in one plant, the other plant may be varying and modifying its genetic make-up to become more lethal or bypass your resistance factors. As everyone knows about the poor old tomato, every time they get a resistant strain, along comes another virus that gets to it and you get wilt again. It is a constant rat race of who is in front of whom in this problem.
        Breeding for disease resistance poses a problem: Can variability of the host to produce resistance to the disease overcome that of the pathogen or parasite. In other words-what we are trying to do is to get ahead of the fungus. You would think that the smart way would be what some people have already advocated-find out what plants are already resistant to the disease. But the book says that this is not the way, that plants which are tolerant of the disease can live in the presence of disease and are usually carriers of the disease. You know the history of Typhoid Mary and if you have a plant that is resistant, say 'English Roseum', it can be carrying the disease but not showing the effects of the disease but it sure isn't going to help you genetically either because what you are trying to do is to develop hypersensitivity to the infection and, in that way, the plant will produce a fungus inhibitor. If you have only tolerance, you are not producing this inhibition. You must breed for hypersensitivity to disease. Take the most sensitive plant you can find and try to work with it until you find one that begins to show signs of hanging on longer. In this way you may be producing true genetic resistance.
        The method of entry and the pathogen should be known to help formulate the possible paths of resistance. How does the disease get into the plant? That is die back, rot rot, leaf spot etc. How do these get into the plant? Through the stomata, roots or elsewhere? This is very fundamental knowledge that you must contemplate before you try to solve it genetically.
        All these things I picked up in reading this book that helped me to realize that I had been wrong in my crosses. It would have saved me a lot of time and trouble.
I would only reiterate for the beginning hybridizer these points:

1. Very seldom is a desirable trait determined by only one chromosome or allele, and neither does a single gene have only one determining function. The whole spectrum of good and bad points must be considered in choosing parents.
2. Selfing often results in sterile seeds.
3. Always use as one parent a plant which has grown and bloomed satisfactorily in your area for at least five years.
4. Water-conserving leaves and disease-resistant roots are probably the two most desirable traits in rhododendron hybridizing. Recent determinations have revealed plants with "C4"" leaves, which are capable of producing more carbon dioxide on less photosynthesis than the conventional "C3" types. Perhaps there is a "C4'" leaved rhododendron out there somewhere.
        In closing, I'd like to say, after it is all over, you are still going to have to throw away. It is only once in a lifetime that you come up with an Einstein in human beings and you'll just have to live with the low odds in rhododendrons hybridizing. I hope I haven't discouraged people. I hope you'll keep at it.


Volume 33, Number 2
Spring 1979

DLA Ejournal Home | QBARS Home | Table of Contents for this issue | Search JARS and other ejournals