JARS 42n4 - How The Japanese Pontica Rhododendrons Fit Together

How The Japanese Pontica Rhododendrons Fit Together
Frank Doleshy
Seattle, Washington

Since Dr. David Chamberlain is headquartered at Edinburgh, Scotland and I in the Seattle area, we communicate mainly by mail - a slow process. Therefore we had to work together more than three years to write a joint paper which is aimed at settling the international disputes about the classification of Japanese Pontica rhododendrons and, more basically, explaining why these plants are in the geographical locations where we now find them. That paper appeared in The Journal of Japanese Botany for August, 1987 - in English with Japanese summary. It reflects Chamberlain's knowledge of rhododendron taxonomy as well as the field work done by myself and wife Catherine (Kay) during our six trips to Japan. We are sure we haven't said the last word on the subject, but hope we've given a fair statement of present knowledge and opinion.
The ARS Journal could not reasonably provide enough space to reprint our paper. To take its place, the editor and I agreed that I should write this shorter article, telling what Chamberlain and I did and how we went about doing it.
I first met Chamberlain in 1978, just after he had let it be known that the botanists might have to take the name "japonicum" away from the familiar azalea and, instead, give this name to the group of rhododendrons we had known as "metternichii" . Numerous ARS members objected angrily to this suggestion, even if they didn't understand why it had been made. Therefore I'm using the rest of this paragraph for a skeleton outline of the rules that botanists follow when naming or renaming plants. Generally speaking, the name they use is the earliest "legitimate" one that has been "validly and effectively" published. "Legitimate" names are those conforming to certain practical rules that are followed by botanists worldwide, for example, that a plant cannot be renamed if it already has an acceptable name, and that identical names cannot be given to two different plants. "Valid publication" requires that the name be accompanied by a description of the plant (or a cross-reference to a description). "Effective publication" has, in recent years, required inclusion in a book or periodical that goes to numerous botanical centers the world over. These various rules, however, can be bent - which happens when the appropriate international committee of botanists decides there is good reason to "conserve" a name even though it does not meet the established standards. For seed bearing plants, the ultimate referee in disputed cases is the International Committee for Spermatophyta (the "Nomenclature Committee"), which publishes its decisions in the quarterly journal called Taxon. The various lists of names that have been proposed for Japanese Ponticas are tabulated on the chart opposite. Those we now use are the ones headed "Hara 1986".

Names Proposed For Japanese Ponticas
Stevenson et al. 1947	Chamberlain, 1982	Yamazaki, 1964	Hara, 1986
R. degronianum	R. japonicum	R. metternichii	R. degronianum
	var. pentamerum	subsp. pentamerum	subsp. degronianum
		subsp. metternichii	subsp. heptamerum
R. metternichii	var. japonicum	var. metternichii	var. heptamerum
		var. hondoense	var. hondoense
		var. kyomaruense	var. kyomaruense
		f. amagianum (added 1981)	f. amagianum
R. yakushimanum	R. yakushimanum	subsp. yakushimanum	subsp. yakushimanum
	subsp. yakushimanum	var. yakushimanum	var. yakushimanum
		var. intermedium	var. intermedium
R. makinoi	subsp. makinoi	R. makinoi	R. makinoi

While the dispute over names was gathering steam, I also discovered that Chamberlain and I did not agree on the classification of Japanese Pontica rhododendrons. This matter of classification is not the same as naming; instead, classification is the grouping of plants into varieties, subspecies, species, and the other categories. Unless a consistent classification is used as a basis for naming the plants, we will probably have superfluous names for some plants and won't have any names for others. Kay and I were more concerned about the classification problems than the name problems, since our interests had turned in that direction during our first trip to Japan, in 1965. On October 19th, we climbed one of the rhododendron-clad Yakushima mountain summits, then we went 600 miles north to cool, foggy Mt. Kusatsu-Shirane, and there on October 23rd found thickets of subsp. degronianum which could almost have passed for subsp. yakushimanum . Yet, between these two places, hundreds of miles of mountain ranges are covered with more lanky rhododendrons which have thinner indumentum. Becoming interested, we looked into the classification of the various rhododendrons, then went still further and tried to figure out how the plants got to their present locations and where they came from.
We'd been happy with the 1964 Yamazaki classification plan (reflected in third column of names on chart), which appeared to be based on natural relationships. Therefore we were startled by the much different Chamberlain proposal published in 1982. (See second column of names.) I explained our objections in the ARS Journal (Vol. 37:2, Spring 1983) and I supplemented this with a long letter to Chamberlain on May 13, 1983. Principally, I criticized the following:
- Treatment of R. makinoi as a subspecies of R. yakushimanum .
- Treatment of var. hondoense as a doubtfully distinct plant.
- Treatment of var. kyomaruense as, probably, a local variant of subsp. pentamerum .
- Recognition of two subspecies within R. brachycarpum .
-Difficulties encountered when we tried to key out our own dried, wild-collected, specimens in accordance with Chamberlain's plan.
Chamberlain replied in a friendly way on June 21, 1983, thanking me for the results of our field observations, and he and I started trying to reconcile our differences. We continued to refer to the "metternichii" group - aware that this name was no longer considered legitimate, yet not knowing whether it would eventually be replaced by "japonicum" or "degronianum" . This, however, did not keep us from trying various rearrangements within the group. He suggested that R. yakushimanum and R. makinoi might be treated as two separate species, with the other members of the group split up into subsp. metternichii , subsp. hondoense , and subsp. pentamerum , and with var. kyomaruense as an intermediate. Although this was a major step, we still had to travel a long road before we arrived at full agreement. The important thing is that we have remained friends, and otherwise would surely have failed.
By going to Japan later that year, Kay and I found ample evidence to support the treatment of R. makinoi as a separate species. Also we looked at various stands of R. brachycarpum and found no significant relationship between geographical location and the amount of indumentum - thus demolishing the idea that this plant should be separated into two species or subspecies on the basis of indumentum. (This was our second autumn trip to Japan with terrible weather, which showed us that Japanese people who talk about the "high skies" of autumn are no more to be trusted than Seattleites talking about Seattle weather.)
Classification of the other Japanese Ponticas is not so easy, and we were inclined to base our suggestions on our ideas about plant origins or migrations - which can be summarized as follows:
First, that the more ancient rhododendrons of Japan had 5-lobed flowers and heavy indumentum, and that these plants survive today as the northern subsp. degronianum (" pentamerum " until recently) and the far-southern subsp. yakushimanum ; also R. makinoi is closely related.
Second, that these were disrupted by arrival of a moderate-size land mass that moved out from the Asiatic coast and now constitutes much of southwestern Japan. (On the map, this is the territory to the left of the Fossa Magna and above the Median Tectonic Line.)
Third, that the new land carried a rhododendron with 7-part flowers and thin indumentum which survives practically unchanged within the boundaries of this land, as var. hondoense .
Fourth, that the newly-arrived rhododendron hybridized with the more ancient rhododendrons to produce present-day var. heptamerum ( "metternichii" until recently) in Kyushu and Shikoku, with numerous intermediate stands to be found in Shikoku and the Kii Peninsula; but in Kyushu these intermediates were wiped out by volcanic activity and replaced by ordinary var. heptamerum . (For detailed explanation see our paper in the Journal of Japanese Botany ; also I gave a less refined but usable explanation in the ARS Journal ; Vol. 37:2, Spring 1983).
These ideas were largely the result of field observations during the trips that Kay and I have taken to Japan. Working from these ideas, and with Chamberlain providing the professional skills and knowledge, we were ultimately able to support the classification plan developed by mainline Japanese botanists. But this required a lot of debate and compromise - especially at a later stage of the game, when Chamberlain suggested major changes in the Japanese plan.
Meanwhile, having returned from our 1983 trip to Japan, we found a letter from Chamberlain saying that the "metternichii" group did not yet have its new name. Then followed a hiatus of several months, broken by a letter from Dr. Herman Sleumer - 78 years old but, after illness, working hard again at the Rijksherbarium, Leiden. This reminded me that we wanted our field work in Japan to result in some progress toward international agreement on classification, rather than merely slipping into oblivion. Therefore I wrote Sleumer and Chamberlain, saying we were considering the possibility of visiting each of them for discussion of the classification problems, then continuing on around the world to see if we could get concurrence from Japanese classification people. Or alternatively we could start the trip in Japan then continue to Holland, and Scotland.
Chamberlain replied on August 2, 1984, saying that our visit to Edinburgh could be most useful, and he suggested that he and I could perhaps plan a joint paper on the Japanese Ponticas. I replied October 2nd, endorsing the proposed collaboration. Then Kay and I left for New England to visit relatives, and from Boston Airport we first drove to Harvard for a talk with Dr. Peter Stevens, head of the Herbaria. He is from Edinburgh, of course knows Chamberlain; and, since Chamberlain and I were already well along toward agreement, he urged that we go ahead on the joint paper. He felt, as I did, that it wouldn't hurt to have our opinions reviewed by a Japanese taxonomist, and he suggested I write the old tiger we all respected - Professor Hara at Tokyo University. He asked a staff member to come in and figure out what to say in the letter, and I quickly sent it. Hara replied on October 29th and surprised me; instead of nominating a junior colleague to review our work, he said he'd by happy to do this himself.
With this correspondence, our idea of going on from Europe to Japan was no longer in the immediate picture. Also, we had thought that old University of Washington friends might join us in Singapore if we took a stopover there (thus keeping a tentative date made 46 years earlier) - but they couldn't see their way to doing this, or maybe they were feeling a little old that day.
I started drafting portions of the joint paper. And, since Chamberlain had been invited to Seattle for meetings of rhododendron groups, he was able to be at our house and talk about the job almost all of Sunday, April 28, 1985. The following Tuesday, the Suzukis visited and gave us a lot of encouragement. Then after express-mailing great bundles of dried material to Leiden and Edinburgh, we flew to Amsterdam, found the train for Oegstgeest (near Leiden) where the Sleumers had made a hotel reservation for us, and had a sound sleep.
We were with Sleumer the next day, mainly at the Rijksherbarium. He had picked out interesting things from their collections, and we saw the earliest dried rhododendron material to reach Europe from Japan. The item most important to us was a specimen which had become a puzzle. This had been collected or obtained by Siebold to represent his new R. metternichii . It was dated 1829, with no source location given. After searchingly examining this specimen (at Sleumer's urging) we felt that something was amiss. Siebold had stated that R. metternichii was a plant of northern Japan, and it has since been assumed that the specimen we were examining was from the north (Nitzelius 1961, 152). Yet Maximowicz, the Russian botanist, seemed doubtful about a northern origin (1870, 21);and Kay and I - after much mental comparison with plants we'd seen in the wild - decided the specimen was almost certainly of southern origin. Perhaps it was picked from a plant which had been transported from the south to the Nikko temples, in northern Japan (since Siebold mentioned direct or indirect contact with the Satsuma lord - from the south). If we thus take it as the southern var. heptamerum (formerly "metternichii" , it is no longer a source of confusion when trying to distinguish the northern and southern Pontica rhododendrons.
Late in the day, we were surprised to learn of a solution to the much-debated problem of deciding which rhododendron should be called "japonicum" . Sleumer showed us his April 18, 1985, letter to Chamberlain, saying that Sleumer's own paper in a German journal (1939, 96) had gone into this matter, with the conclusion the " japonicum " should be used for the azalea. Kay and I did not see Chamberlain when we went on to Scotland (since he had left for China) and we don't know what transpired after that, but we later learned that Hara, as a Nomenclature Committee member, had strongly advocated that " japonicum " remain the name of the azalea. On November 13th, Chamberlain replied to Sleumer, acknowledging the importance of Sleumer's 1939 paper, and saying the Committee had indeed reached the same conclusion as Sleumer. When I read the Committee's report on this, in the February, 1987, issue of Taxon , I saw that they referred to the same source as Sleumer had.
Also, on November 11, 1985, just before he wrote Sleumer, Chamberlain phoned to tell me of the decision to keep " japonicum " for the azalea. Since he, more than I, was concerned with naming, this was a major "go" signal for him, allowing him to proceed on his part of our joint paper without so much uncertainty. On December 11th, I sent him a new draft of the paper but soon received his letter of December 23rd proposing that a large part of the Yamazaki classification be changed as follows:

Yamazaki 1964		Chamberlain Dec 23, 1985
R. metternichii subsp. pentamerum		R. degronianum subsp. degronianum
		var. degronianum
		var. kyomaruense
subsp. metternichii		subsp. metternichii
var. metternichii		---
var. hondoense		---
var. kyomaruense		---
		subsp. hondoense

Renaming of R. metternichii to R. degronianum was a matter of housekeeping, reflecting the fact that the old name had not been legitimate, therefore needed to be replaced. The other changes were proposed reclassifications. First, Chamberlain wanted to move var. kyomaruense to a new spot under subsp. degronianum ; second, he wanted to redefine subsp. metternichii so that it included only the plants that had been in var. metternichii , and then elevate var. hondoense to the subspecies level. His proposals did not break the technical rules for plant naming, since these rules would allow " metternichii " to be used as a subspecies name, even though not a good species name. Consequently, if he had succeeded with his proposals, the " metternichii " name would still have been available for use by those who prefer old names to new.
But the big question was whether or not the proposed reclassifications reflected the actual relationships between the plants. To me, the answer had to be "no", after beating my way through thousands of these plants, looking for geographical gaps and for differences in plant characteristics. In detail (and using the Yamazaki names) here is why I could not agree with the proposals: - I had seen no convincing evidence that var. kyomaruense and subsp. pentamerum belong together under one subspecies, even though they both have 5-part flowers. They are found on opposite sides of the Fossa Magna, the immense depressed area at the northeast end of the land thought to have moved from the Asian coast. This is a barrier that may have varied in width. But var. kyomaruense , except for its 5-part flowers, is almost indistinguishable from var. hondoense , so is not just a displaced pentamerum . And I do not agree that the 5-part trait crossed the Fossa Magna, got into the 7-part var. hondoense , and thus produced var. kyomaruense , because:
(a) Hybridizers working with Japanese Ponticas have found 7-part flowers dominant over 5-part; thus a 5-part-to-7-part invasion would appear to be upstream.
(b) I don't think a recessive 5-part trait could get from subsp. pentamerum to var. kyomaruense without bringing along other distinctive traits of the pentamerum . Instead, I think the 5-part character of var. kyomaruense arose independently within the kyomaruense-hondoense complex. - Now looking at the relationships within the var. kyomaruense-hondoense-metternichii group, vars. kyomaruense and hondoense are geographically separated by numerous north-south ridges and valleys which do not provide easy east-west access for plant genes (or people!), but this cannot be considered a very penetration-proof barrier in the long run. Vars. hondoense and metternichii are not even separated by that much of a barrier; instead they meet and merge in a zone extending through much of Shikoku and the Kii Peninsula. (See map.) Not one of these three plants is nearly as distinct as subsp. pentamerum , which is separated from them by the Fossa Magna and is distinct from them because of its similarities to subsp. yakushimanum . Therefore I think our classification would be less realistic and useful if var. hondoense were elevated to a subspecies.

I'd like to have found some way to bridge the gap between Chamberlain's proposals and the conclusions Kay and I had reached from looking at the wild plants. But I could not, and the time had come to get the views of a Japanese botanist. Therefore, letting Chamberlain know, I incorporated his proposals in a draft of our paper and sent this to Hara for review on May 6,1986. Hara replied promptly, saying he'd read the draft with great interest but that we were a little late. After settlement of the question about the " japonicum " name, it had become urgent for Japanese botanists to get their Pontica rhododendrons in order; therefore he had already done so in a short paper to appear in the Journal of Japanese Botany for the following August. When I turned to the outline of his new arrangement (col. 4 of name table), I saw he had brought in the necessary new names, such as R. degronianum for R. metternichii , but had not accepted Chamberlain's proposed shift of var. kyomaruense or proposed realignments. Instead, he had stayed with Yamazaki's classification plan, which had proved excellent in the field - and this suited me right down to the shoelaces. He added that our paper included "interesting and useful observations and opinion", and he suggested publication in the Journal of Japanese Botany if we wished. This of course, would fulfill my long-held hope of having a Briton and American speak together with a strong voice in Japan.
I phoned Chamberlain immediately (June 3rd) and followed this with a letter on June 9th. Also I wrote Hara on the 13th, expressing regret for the slow progress Chamberlain and I had made but telling him of my strong agreement with his arrangement of the Ponticas, as well as my pleasure with his suggestion of publication in the Journal of Japanese Botany.
Chamberlain responded with a letter of July 4th, mentioning his regret at loss of the " metternichii " name (at subspecies level), and suggesting that we needed to determine whether the Hara classification best expressed the geographical distribution of the Ponticas. But he said he'd be willing to go ahead and modify our paper to incorporate the Hara classification if we could agree that this involved no compromise of our principal views. I replied at length, explaining that the Hara classification was indeed compatible with my field observations, and I told him I thought the use of the classification would improve our paper. I agreed that we should see the complete Hara paper, and I wrote Hara, asking that a copy be sent me as soon as available.
Then, stimulated by the opportunity to work with a classification plan that did not trouble me, I went ahead with necessary revisions in the paper. Beyond this, I did a comprehensive rewrite with numerous rearrangements. On nearly every page there were things to be expressed more clearly or crisply, and the job was now a pleasure.
A printed copy of Hara's paper arrived September 22nd - in Japanese as I expected. Therefore, not being very confident that Chamberlain could get speedy and accurate translation service in Edinburgh, I did the job. Since I'm not fast at this myself, I didn't finish until October 2nd, when I mailed it to Chamberlain. Also, to get any criticism of this job, I sent a copy to Hara the same night. But he had died, on September 24th. I learned of this tragedy in a telephone conversation with Chamberlain October 22nd. This, as usual, was at 11:15PM here, 7:15 AM there, and we quickly became very hushed. I assured him that Hara's paper had indeed been published, and he told me he'd have a revised manuscript off to me within a week. We wondered who we could turn to, now, at the University of Tokyo. But we needn't have worried; Professor Ohba wrote on October 23rd, letting us know that our paper was still in good hands.
After that, the job was simply a matter of final refinements, and I mailed our final version to Ohba on January 23, 1987. On March 23rd, he replied that it was accepted, and it appeared in the August Journal of Japanese Botany . Both the Chamberlains visited us a few weeks later, after having had garden tours galore. I took them to the UW campus where, long ago, a math professor had got me interested in spatial relationships - therefore happy when working with any geographical problem, including the origins of plants. They looked around with interest, noticing the similarity to the Oxford campus, admiring the clean-cut kids, and finally wondering if this was the place for their sons.

References and suggestions for additional reading
Chamberlain, D. F. 1982. Revision of Rhododendron II. Subgenus Hymenanthes. Notes RBG Edinb. 39(2):209-486.
Chamberlain, D. F., and F. L. Doleshy. 1987. Japanese members of Rhododendron subsection Pontica: distribution and classification. (English with Japanese summary) J. Jap. Bot. 62:225-43
Doleshy, F. L. 1983. Distribution and classification of certain Japanese rhododendrons. J. Amer. Rhod. Soc. 37:2 81-89.
Doleshy, F. L. 1984. The north half of Japan, 1983. J. Amer. Rhod. Soc. 38:2-5 29-31.
Du Rietz, G. E. 1930. The fundamental units of biological taxonomy. (English) Svensk Bot. Tidskrft. 24:333-428.
Geological Survey of Japan. 1968 and later eds. Geological Map of Japan, Volcanoes of Japan, Tectonic Map of Japan. (Japanese& English)
Scales from 1:500,000 to 1:5,000,000. Pub. at Kawasaki by the Geolog. Survey.
____________ 1982 ed. 1:500,000 Geological Map, Sheet 77, Kyoto. (Japanese & English). Pub. as above.
Hara, H. 1986. Scientific names of the Rhododendron degronianum group (Japanese with summary in Latin & English). J. Jap. Bot. 61:245-7.
Maximowicz, C.J. 1870. Rhododendreae Asiae orientalis (Latin). Mem. Acad. Sci. St.-Petersbourg . Ser. 7, Tome 16, 9:1-53.
Maekawa, F. 1974. Geographical background to Japan's flora and vegetation: General geography of Japan and its relationship to the flora, pp. 2-20. Also, Origin and characteristics of Japan's flora, pp. 33-86. In The flora and vegetation of Japan (English). M. Numata, ed. Tokyo: Kodansha.
Nitzelius, T.G. 1961. Notes on some Japanese species of the genus Rhododendron (English) Acta. Horti Gotoburg . 24:135-174.
Sleumer, H. 1939. Kritische Bemerkungen zu K.A. Reiter "Azalea mollis und Azalea mollis" (German), Jahrb. Deutsch. Rhod. Ges. , Bremen, pp. 95-96
Stevenson, J.B., ed 1947. The species of Rhododendron. 2nd ed. London: The Rhod. Soc.
Takai, F., T. Matsumoto, and R. Tonyama, eds. 1963. Geology of Japan (English). Tokyo: Univ. of Tokyo Press.
Yamazaki, T. 1964. A new variety of Rhododendron metternichii and its alliance (Japanese with summary in Latin & English). J. Jap. Bot. 39:13-18.

Frank Doleshy and his wife Kay, Seattle Chapter members, have traveled extensively in Japan tracking the Japanese Pontica rhododendrons.