JARS 42n4 - The Nature Of Rhododendron yakushimanum Clones

The Nature Of Rhododendron yakushimanum Clones
M.J. Harvey
Halifax, Nova Scotia, Canada

This report is an offshoot of a continuing breeding program which had the aim of producing compact, indumented Rhododendron hybrids with excellent or superb foliage. Some of the ideas behind this project have already been laid out in an article: "Forget The Flowers - Breed For Leaves," American Rhododendron Society Journal (Vol. 39:3, Summer 1985).
Now when you start a breeding program the general principle is to obtain the very best forms as the parental stock. For this reason I have assembled over the past few years a group of indumented species from friends, commercial sources and the Rhododendron Species Foundation. Most of these young plants have not flowered yet but by the use of more mature plants in friends' gardens and pollen from various sources I have been able to produce a number of F1 hybrids already.
Only two indumented plants of mine have flowered and it is these, perversely, which have caused problems. They are two forms of R. yakushimanum (yak), the clone 'Ken Janeck' and a Leach seedling which we have locally dubbed 'Sister of Mist Maiden' since it came from the same batch of seedlings from which 'Mist Maiden' was released. It is probably the seedling which was later named 'Pink Parasol' but not necessarily so, hence I have retained our informal name.
I was initially keen to obtain plants of these more robust forms since their vigour, larger flowers and deeper pink colouration make them extremely desirable plants and potentially good breeding stock. These vigorous forms are in fact best represented by the aforementioned clones 'Ken Janeck', 'Mist Maiden', and 'Pink Parasol'.
However, no sooner had I flowered, cross-pollinated and distributed seed from my yaks than people started whispering suggestions in my ear that these yaks were not really pure - that they were, in some degree, of hybrid origin.
Now this was rather upsetting and I did not really want to hear such talk. I had two reasons for not wanting to believe this rumour. First, the plants I have really are magnificent and just what one has in mind as the ideal breeding stock. Second, if true I would have to start again and obtain new plants. But the dilemma here was, if I did get new plants, how could I be certain that they were not also hybrids. I had visions of packing a spade and setting off for Yakushima, just to be sure.
The rumour may not be true, in which case I could continue my breeding program. But this all left me in a very unsatisfactory frame of mind and it became obvious that I needed a test to tell whether a given yak was the pure species or a hybrid.
In a situation like the above the first resort is to go to the literature. A perusal of various books revealed very little in print. This is not all that surprising since, quite rightly, people are reluctant to commit vague feelings to paper when they have no concrete evidence. There is a danger of committing a sort of horticultural slander. Cox, in The Smaller Rhododendrons (Batsford, 1985) says of R. yakushimanum , "There are many selected and several named clones. Some of the latter may in fact be hybrids and have been raised from seed off cultivated plants...introduced to Exbury, England in 1934 (two plants from Koichiro Wada)...'Ken Janeck' Award of Excellence, is a stronger-than-average grower with unusually flat pink flowers. Probably a hybrid. 'Mist Maiden' large flowers heavily flushed-pink on a strong plant. 'Pink Parasol' larger leaves and deeper pink flowers than average."
So Cox suspects 'Ken Janeck' to be a hybrid and leaves open the situation in the case of 'Mist Maiden' and 'Pink Parasol'. He does however provide a mechanism by which a hybrid state could have inadvertently been achieved: the raising of seed from open-pollinated plants situated in gardens where they are surrounded by other species and hybrids.
I see that Polly Hill, in a recent article "Three New Yaks From Barnard's Inn Farm", American Rhododendron Society Journal (Vol. 41:3, Summer 1987), has taken just this attitude, that is, seed that comes in a packet with a label will not necessarily produce seedlings corresponding to that label. Of the three clones she describes she states that the seed which produced 'Wild Wealth' was collected on Mount Miyanoura of Yakushima in 1961. It is in fact a reasonably vigorous example of the species.
The other two yaks described by Polly Hill ('Samisen' and 'Big Yak') are stated by her to be definitely F1 hybrids but to have originated from seed described (embarrassingly) as "self hand-pollinated Exbury Form at Washington" in the ARS seed list for 1966. Ah, the problems of beating the bees! I actually wrote a short article on pollination for the Rhododendron Society of Canada entitled "How To Make Your Rhododendrons Cross" ( Bull. RSC 16:14-17,1987). In the article I described the use of "bee cheaters" which largely avoid the problems that unwanted pollen brings.
Refreshing as it was to see Polly Hill's attitude it still left me with a problem regarding the currently available plants. Basically I needed answers to the following nested set of questions:
1.  Are some R. yakushimanum clones of hybrid origin?
2.  If some are hybrid, which are and which are not?
3.  Is there a simple test to determine hybridity?
I should, first of all, report on which specimens were examined in spring of 1987. None of my own yaks flowered in 1987 so I am indebted to various friends for permission to visit their gardens to obtain material. I thank Tom Baskett, John Bond, Walter Ostrum, Dick Steele and John Weagle for donating the pollen and flowers that I used. The following list is not a systematic survey but merely the plants which happened to be available to me at the time.

Plants examined
1.  'Koichiro Wada' (FCC form) (KW), one of the three original plants sent by Wada to Lionel de Rothschild at Exbury in 1934 (two survived).
2.  Collected Wild seedling (CW), one of the seedlings resulting from a more recent collection from Yakushima.
3.  'Koichiro Wada' x Collected Wild (KWCW), a hybrid made between the two above.
4.  WCP84376 (WGP), one of the clones in Windsor Great Park, England.
5.  'Ken Janeck' (KJ), the strong-growing USA West Coast cultivar given the Award of Excellence.
6.  'Mist Maiden' (MM), another vigorous cultivar (D. Leach).
7.  'Sister of Mist Maiden' (SO), sibling of the above, similar strong form (probably 'Pink Parasol' but the label lost).
8.  Gable's smir-yak (SY), a hybrid of R. smirnowii and R. yakushimanum produced by Joe Gable.
The choice of how to test for evidence of hybridity then came into consideration. There is now available a tremendous range of methods that can be used. Especially valuable are the many biochemical techniques. Examples of these include the isolation and characterization of marker substances such as flavonoids, terpenes, sterols, alkaloids, isozymes and DNA. The techniques include thin-layer and gas chromatography, gel electrophoresis and a variety of DNA analysis methods. Using these we could probably get evidence as to the ancestry of, for example, the Ironclad hybrids.
However, being in the increasingly common position of many scientists these days of having no money for research, I was in the position of being free to use any technique I desired - just so long as it cost less than $20. I chose to use a ruler and a microscope since both had been purchased over 20 years ago. This restriction cuts out any modern sophistication but it does have the advantage of enabling anyone to repeat my method.
The best evidence came from flower size and colour, and from pollen grain morphology. The flowers I sampled in a rather unscientific way by just removing representative flowers and photographing them, Fig 1. I did not want to ruin the flower displays of my friends by removing large numbers of blooms. The latter would have been the preferred statistical method but I hesitated to try the patience of my colleagues any more than I had. Hence you will have to take my word that the flowers in the photograph were typical examples of their clone.

R. yakushimanum flowers from 
different clones
Figure1. R. yakushimanum flowers, rear view. Left to right: 'Mist Maiden',
'Koichiro Wada', 'Koichiro Wada' x Collected Wild, Collected Wild.
Photo by M.J. Harvey

Pollen as an indicator of hybridity
The pollen evidence requires a little explanation because of the unusual nature of Rhododendron pollen. During the formation of pollen a pollen mother-cell divides by meiosis to give a cluster of four pollen grains. (Meiosis is the pair of cell divisions which reduce the number of chromosomes in each cell to the gametic number, n). In Ericaceae, unlike most plant families, these sets of four grains adhere tightly together in a symmetrical cluster in such a way that the centre of each cell is at the apex of an imaginary regular tetrahedron, Fig 2a. In addition, these clusters of four are attached together in chains by threads of an elastic material. This latter need not concern us here as the elastic material was partly dissolved when the pollen was stained.
The rationale for examining pollen grains is that during the division processes that lead to the formation of pollen, the chromosomes must pair in a very precise one-to-one relationship. If the plant is a hybrid between two distant or moderately distant species, then the chromosomes may differ in structure to some degree and pairing failures will occur. The subsequent maturation of the pollen cells is very sensitive to chromosome irregularities, in fact the loss or gain of even a single chromosome can cause the grain to be inviable. Hence, working backwards, the presence of empty or malformed pollen grains can be used to indicate meiotic chromosome irregularities and, by extension, lead to the deduction that such a plant is of a hybrid nature, Fig 2, b, c, d.

Rhododendron pollen tetrads
Figure 2. Diagrammatic representation of Rhododendron
pollen tetrads. A. Four normally formed grains with centres
at the apices of a regular tetrahedron, with the grain out
of the plane of the paper represented by a circle
(omitted from remaining diagrams); B., C., D., one,
two or three grains of the tetrad malformed and empty.

Why not, one might inquire, look directly at the chromosomes during the pairing process instead of several steps down the line at what happens when something goes wrong? You can in fact do this but the problem here is that Rhododendron chromosomes are small and it is difficult to catch them at the right stage of meiosis. Apparently there are just a few days in the year during which the young flower buds are at the right stage. Even so it is surprising that so little work has been done on Rhododendron chromosomes. There was one person, Janaki Ammal, who in the 1940's performed an amazing amount of pioneer work on Rhododendron chromosome numbers. Since then almost nothing has been added to her results. It is high time that this line of research were extended. In contrast to the chromosomes, the pollen grains are easy to observe and photograph, are preserved by drying and need no critical timing in collection.
In plant families where the pollen grains separate at maturity the normal procedure is to count the good and bad grains in a sample and report the bad grains as a percentage. The tetrad clusters in Rhododendron do not permit this method of scoring because it is not possible to examine each individual grain clearly. Fortunately, in the case of the samples under consideration here it was not necessary to use statistical subtleties since it can be seen from the photographs, Fig 3, that the samples fall into either of two categories: all good, or with a substantial number of hollow or shrunken grains in the tetrads.

R. yakushimanum and Gable's Smir-Yak 
pollen grains
Figure 3. Photomicrographs of R. yakushimanum and Gable's Smir-Yak pollen grains.

Among the vegetative characters Fig 4 illustrates the extremes of size by contrasting two, three-year-old rooted cuttings; the smaller 'Exbury' clone with small leaves and short shoots, with the larger and flatter-leaved 'Ken Janeck' showing vigorous shoot extension. The leaves themselves show great variation as can be seen in Fig 5 where the longer and broader leaves of 'Ken Janeck' and 'Sister of Mist Maiden' contrast with the long but very narrow shape and re-curved margins of 'Yaku Angel' and the short, re-curved leaves typical of the more compact forms such as 'Exbury' and 'Koichiro Wada'.

'Ken Janeck' and R. yakushimanum 
Exbury comparison.
Figure 4. Rooted cuttings approximately three years
old of 'Ken Janeck' and R. yakushimanum Exbury.
Photo by M.J. Harvey
'Underside of leaves comparison.
Figure 5. Underside of leaves. Left to right,
'Ken Janeck', 'Sister of Mist Maiden', 'Yaku Angel'
and R. yakushimanum Exbury.
Photo by M.J. Harvey

Discussion
It can be seen that the plants under consideration fall into two groups. The first group includes those that are characterized by relatively robust growth, larger leaves, deeper and more persistent pink blush to the flowers, and the presence of defective pollen grains. To this can be added thinner fruiting capsules containing fewer seeds when pollinated with the same pollen as the other group.
The second group consists of plants exhibiting a generally dwarfer growth habit (although there is a range here), their smaller flowers bleach to white by the time they are fully expanded, they produce uniformly good pollen and have fatter capsules. These results are summarized in Table 1.

Table 1
Table 1

It is the thesis of this article that the robust group, which includes 'Mist Maiden', 'Ken Janeck', 'Sister of Mist Maiden' and 'Pink Parasol' represents a series of hybrids resulting from stray pollen landing on the stigmas of the yak plants which were the source of the seeds used to originate these clones. In contrast are the generally more compact specimens, with the whiter flowers and good pollen which I maintain represent various forms within the pure species. These include 'Exbury', 'Koichiro Wada', Collected Wild, the hybrid between the latter two and WGP 84376. I have not had the opportunity to examine pollen of 'Yaku Angel' but from its narrow leaves and whitish flowers I would surmise that it represents a specimen of the species.
The suspicions about the existence of hybrids have circulated for some time, the clinching evidence brought forward here is the existence of the two pollen groups which give a definite sorting of the plants into two classes, correlating nicely with the more familiar evidence of growth and colour.
As luck would have it a plant was available of the hybrid of R. smirnowii and R. yakushimanum made by Joe Gable. In growth and flower characteristics this plant is virtually indistinguishable from the robust, more colourful plants discussed above. It is significant that its pollen also shows defects to about the same extent as the other presumed hybrids. This is strong support for the hybrid hypothesis but it should not be interpreted to mean that the robust plants are all hybrids of R. smirnowii . There may have been more than one species or hybrid contributing pollen to the R. yakushimanum plants used as the seed parents. 'Ken Janeck' for instance apparently had an independent origin from the Leach plants. We need more evidence as to the probable nature of the male parent or parents before we can come up with an answer. Whatever it or they were a strongly indumented leaf would appear to be one of their conspicuous characteristics.
It would be nice to trace back the origin of the various plants involved and I am indebted to David G. Leach for the following information with respect to his plants. The seed was obtained from Francis Hanger who at that time was in charge of the gardens at Exbury. It was the result of crossing the two surviving plants of the three yaks which had been sent from Japan by Koichiro Wada. At the time these were the only two plants of yak in the western world. They are very similar and represent the more dwarf and slow-growing forms of the species. People have puzzled for years whether they are members of the same clone. Opinion seems to be that they are genetically distinct, they appear to set more seed when crossed than when selfed and this is evidence of genetic separateness. One is known as the 'Exbury' form and the other is now known as 'Koichiro Wada'.
From the seed received by Leach some 400 plants were raised all of which were indumented and much alike. From these Leach selected seven which had an edge in flower quality and in hardiness in the Appalachian foothills at Brookville, Pennsylvania. From this group of seven were finally selected 'Mist Maiden' and 'Pink Parasol'. The one informally named 'Sister of Mist Maiden' was one of a number of cuttings derived from these seven and given to R.M. Steele. My interpretation of the above is that of the seeds received by Leach most represented pure yak, being the cross 'Koichiro Wada' x 'Exbury', but that mixed in with them were a very few seeds resulting from pollen of some other plant getting on to the stigma of a flower and fertilizing a few ovules. These few hybrids attracted attention to themselves when planted out by their more robust growth and deeper colour and were eventually chosen as the superior forms.
I should conclude by saying that in dividing the currently available yaks into two categories: hybrids and the pure species, I am not in any way intending to disparage the hybrid group. Indeed, from the point of view of ease of growing and sheer spectacle, both in leaf and flower, the hybrids are preferable. My only intent is to warn hybridizers of a possible unknown parentage factor when they use representatives of the hybrid group as breeding stock.
I thank David Leach for an excellent lunch at no notice for my wife and myself and for a wide-ranging discussion about the whole yak problem. He does however reserve the right to interpret the observations in his own way.

Joe Harvey is a biologist at Dalhousie University, Halifax, interested in breeding good foliaged, hardy rhododendrons.