JARS v38n1 - The North Half Of Japan, 1983

The North Half Of Japan, 1983
Frank Doleshy, Seattle, Washington

After 12 years' absence, Mrs. Doleshy and I returned to Japan in 1983, staying from October 5th to November 5th. We started with a two-week visit to the north island of Hokkaido then spent the remainder of our time on Honshu, going first to the mountains near latitude 40 and finally to R. makinoi territory.

Hokkaido
In area, Hokkaido is about the size of South Carolina or western Washington, but it spreads out three-quarters as wide as Oregon and extends north-south through approximately the same latitudes as that state. Since neither of us had been there since 1945, when I arrived via landing ship for a brief visit, we needed help with our plans. For this we consulted Prof. Tsujii, Curator of the Hokkaido University Botanic Garden; he was then kind enough to pick us up at the Sapporo airport after our long day's journey from Seattle via Tokyo, and he made many travel arrangements that smoothed our plan. Two weeks was barely enough time for an introduction to this part of Japan, but by following Prof. Tsujii's advice we saw numerous interesting places, including several that deserve a longer future visit.

From Sapporo we flew to Wakkanai (see map) then west via 12-passenger airplane to Rishiri Island, on which the 5640-foot central peak was already snowy. Walking into the foothills as far as we could during the one afternoon we had there, we did not see any rhododendrons. But the next morning, when we went to the ferry terminal for the return trip to Wakkanai, we looked at a plant seller's offerings and saw R. camtschaticum and R. aureum , which she said came from the central peak. Both these species also grow in the Soviet Far East, 75 miles away, and it would be interesting to determine whether the Rishiri plants are more closely allied with those to the north or those to the south.
Arriving back at Sapporo airport, we picked up our rented car and started the 115-mile drive south to Mt. Apoi, near Cape Erimo. For our visit there, Prof. Tsujii had arranged that we stay in the guest quarters of a forestry lodge, operated much like a regular Japanese inn but charging about half the going rate and providing unusually spacious rooms. When we arrived we were greeted by the forestry staff for that area; they wanted to know how we felt about climbing Mt. Apoi with them the next day, and we assured them that this was what we wanted to do.

The mountain, as we knew, consists of serpentine rock, producing a soil situation also found in parts of the western U.S.A. - where it commonly results in stunted vegetation or barren land. Yet in 1971 we had visited a serpentine area of Honshu and found an unusual brachycarpum -like rhododendron with rounded leaves, a plant which had been difficult to identify without flowers or seed capsules. Therefore we were curious about the rhododendrons we might find on Apoi. And, on this year's trip, we were doing some collecting for University of Washington Prof. A.R. Kruckeberg, who is not only working on Japan-Pacific Northwest plant relationships but also is a major authority on the plant life of serpentines.
Accompanying us the next day, October 11th, were Mr. Nakagami, forestry chief of the Apoi area, Mr. Mitsugi and Mr. Akama from the forestry headquarters a few miles up the coast at Urakawa, and 72-year-old Mr. Onishi, a respected gentleman apparently serving as volunteer guard-conservator. We did not have to go far to see R. brachycarpum , growing at the surprisingly low elevation of 300 feet - and also as we found, on the 2660-foot summit. This plant's height were usually 4 to 7 feet in light shade, 8 to 9 feet in heavy shade. We did not see it in open areas where it would have been entirely unshaded. Despite the unusual rock and soil, both the lightly and heavily shaded plants appeared healthy, with deep green leaves in normal size ranges. But they had two unusual features: first, an indumentum so thick that it was opaque, covering the leaf undersurface except the midrib, and second (especially on plants growing below 1100 feet) a broadly rounded leaf shape.
This R. brachycarpum , as found near 1100 feet, turned out to be practically identical with the previously-mentioned Honshu plant which we'd had trouble identifying in 1971. Comparing our dried specimens of the two, the only difference we can see is a red coloring in the leafstalk and midrib of the Honshu plant. Therefore we now feel sure the latter is R. brachycarpum , but we are still uncertain whether rounded leaves are consistently associated with serpentine, and if so, why.
The R. dauricum and R. kaempferi on Apoi were common in open, shrubby places from about 600 feet upward. Both grew as fairly dense shrubs, 2 to 5 feet tall, but were otherwise very distinct; the R. dauricum already had purplish-bronze shading in its leaves, while the R. kaempferi leaves were still quite green, only starting to turn red, and had characteristic short, sparse hairs on their upper surfaces. We did not see R. albrechtii until we reached an elevation of 1020 feet, where it was growing in fairly dense woods. These plants were 6 to 8 feet tall.
On this small mountain, located at lat. 42°07', long. 143°01', all the life zones seemed to be far lower than usual, with a conspicuous transition to shrubby vegetation at about 1640 to 1800 feet. Then the summit itself, at 2660 feet, had a crest of short but not particularly shrubby birch trees, and R. brachycarpum grew among these, even though not found in the more open zone just below.
In response to my question, Mr. Onishi said R. camtschaticum occurred on the ridge connecting Apoi with the main Hidaka Range to the east. We did not have time to go onto that ridge, and Mr. Onishi looked hard for any stray plants near the Apoi summit but found none.
We collected the following rhododendrons with seed:
No. 802 R. dauricum , elev. 575 feet
No. 803 R. kaempferi , elev. 690 feet
No. 804 R., albrechtii , elev. 1020 feet
No. 810 R. brachycarpum , elev. 1100 feet
Also, from R. brachycarpum without seed, we obtained herbarium specimens as follows:
No. 808 with very rounded leaves, elev. 460 feet
No. 809 elev. 1080feet
No. 811 from summit, 2660 feet
Moreover, for Prof. Kruckeberg, we collected an Aruncus , a rare, low-growing Sorbus (mountain ash) about 8 inches tall, a Potentilla species also found from Alaska to California, and a rare, endemic birch.
With such a variety of interesting plants, the vegetation of Apoi does not resemble the impoverished plant life on some serpentines of Washington and Oregon. Apparently such soils vary in toxicity, perhaps depending on their chemical composition or the speed of weathering.
Late that afternoon, one of the forestry managers at Urakawa invited us to visit for a discussion. He shared our interest in plant geography and classification, and one subject we talked about in some detail was the problem of distinguishing R. dauricum from high-elevation R. mucronulatum . He asked if we were interested in a dwarf, small-leaved R. dauricum found at higher elevations in this part of Hokkaido. We responded with enthusiasm but looked at the map and saw we could not visit any of this plant's habitats without taking a day from one of our subsequent stops. We now wish we'd made this trade; but being unable to foresee the next few days' disappointments, we decided to stay with our schedule. Therefore he arranged for us to stop the next day, October 12th, at the private garden of Mr. and Mrs. Abiko, on the other side of Cape Erimo. This garden turned out to be a beautiful place with many kinds of rhododendrons, and the little R. dauricum was similar to the dwarf R. mucronulatum (as it is called) that we had seen near the summit of Mt. Halla, Korea, a year earlier. Our hosts found all the seed they could, which we took as our No. 812 - to be grown with caution because of the possibility of random crossing in a garden, but likely to produce interesting results.
At the urgent invitation of these hospitable people, we went with them to a large, public rhododendron garden near the Cape (the only such public garden in Japan, they told us), and there saw hundreds or thousands of R. brachycarpum plants, along with large numbers of native azaleas. The R. brachycarpum , they said, was the Cape Erimo form, distinguished by leaves even more rounded than those found on the Mt. Apoi plants. We agree with this as a general statement, although we had seen some individual Apoi plants with leaves just as rounded. Looking at indumentum, we could not find any consistent difference between that on the Apoi and that on the Cape plants.
Reluctantly leaving, we head for our overnight stop at Kushiro, and then found ourselves in the frustrating interval, several days long, when we traveled and climbed in eastern Hokkaido but could not seem to get to any rhododendrons. Our fortunes finally improved October 18th at Onneyu Spa, on the main east-west highway across the central mountains. Prof. Tsujii had suggested a one-night stop there to see a big stand of R. dauricum ; we arrived at our inn by 10:00 a.m., told the desk people that we'd check in later, and were guided by one of the maids to a place where we could see the R. dauricum . This turned out to be a nursery only seven or eight blocks away, dealing in assorted ornamental plants and located at the foot of a ridge where enormous masses of the wild R. dauricum were growing in plain sight. To these latter plants we wanted to go, but a woman at the nursery insisted we could not.
Finding no nearby access to the ridge, we drove slowly back toward a shrine which we thought might have a path to it. Seeing a reasonable-looking couple who had just put their car in the garage, we stopped and asked them about going to the ridge, and they also replied that we could not do this. But we were not ready for another day of failure after the preceding five days; I talked and talked in my kindergarten Japanese - working as hard for twenty minutes as I ever had in my life - and somehow got it across to them that we were interested in the geographical distribution of wild plants, therefore would like to collect some seed and specimens which would give us an unbiased representation of the wild plants, rather than any cultivated plants. They then became very thoughtful and finally seemed to agree that this was a reasonable objective. With warnings about fire danger, they showed us the entrance to a trail leading up the slope, where R. dauricum was the major underbrush.
Individual plants were broad and bulky, 4 to 7 feet fall. Our collection, which included ample seed, was No. 81 6, from an elevation of only 525-625 feet, lat. 43°45', long. 143°3'. Along with this R. dauricum , and primarily for Prof. Kruckeberg, we collected a Ledum (different from one obtained a few days earlier), as well as a maidenhair fern and a very tall, thorny shrub.
We noticed several places where the R. dauricum grew thickly among snags of fire-killed birch, and we later speculated that this rhododendron is present in such large quantities because it moves quickly into openings created by forest fires, but that the rhododendron itself is very flammable and can be temporarily wiped out by a fire. Thus fire may be both friend and enemy, leaving local government in a quandary about the degree of protection desirable.
Light snow flurries came and went throughout our afternoon with the R. dauricum , causing us to wonder how we would fare the next morning when we continued over Sekihoku Pass to the main Daisetsuzan Mountains. But our surefooted little car had no problems on the packed snow. We arrived without incident at Sounkyo Spa and ascended via a ropeway (the Japanese term for any aerial cable car) to the 4200-foot level. Near the uphill station, we saw several plants of R. brachycarpum ; hoping to find more, we went to the chair lift that continues up the slope, but found it was not yet operating. Therefore we walked uphill beneath it to the head house at 5000 feet. Having seen no more rhododendrons along the way, we debated walking still higher - finally deciding this was impractical because the snow was already hip-deep in some places, making it likely that R. brachycarpum - and any R. camtschaticum - would be hidden. Therefore we returned to the 4265 foot level for the R. brachycarpum there - our No. 821 (October 19th, lat. 43°42', long. 142°57'). In comparison with the Apoi plants, this turned out to have leaves less rounded and a thinner indumentum with nearly bare patches.
While slogging down through the snow, we reflected on the desirability of scheduling Hokkaido trips a few weeks earlier in the fall, or at least going first to the central mountains and later to the relatively balmy south coast. These impressions became even stronger the next day, when we went to Yukomanbetsu Spa. Prof. Tsujii had recommended this place because of a two-section ropeway which climbs from the 3600 to the 5230-foot elevation, reaching the zone of alpine flora. As shown on the postcards, the plants include R. camtschaticum , pink heather, large areas of R. aureum , all on a gentle slope leading up to the steam jets in Mt. Ashai's breached crater.
We arrived at our inn by 1:30 p.m., quickly rode up the ropeway, and found one to three feet of snow at the top, with more coming down. All we could do was look for plants in spots blown clear by the wind; we were wondering how long to do this when the ropeway people rang their bell to signal a close-down, and we descended with the entire staff of the uphill stations. Our comfortable inn had no other guests (between climbing and skiing seasons), and we quietly worked on our seeds.
Next afternoon at Sapporo, we regretfully turned in the little car that had served us well. The following day was the first for using our unlimited, first-or-any-class National Railway passes, and we started with a side trip to Otaru, where I had stepped onto the beach from the landing ship for my first visit to Japan.

Northern Honshu
Via an overnight stop at Hakodate and the ferry to Aomori, we crossed over from Hokkaido to Honshu, then took a train to Morioka. Three of the high mountains at Morioka's latitude support the southernmost known stands of R. camtschaticum , and we had planned to climb at least two of them.
6695-foot Mt. lwate is practically in Morioka's suburbs. The 2065-foot level can be reached in about an hour by bus and taxi, and from there a good trail leads up through azaleas to R. brachycarpum at about 3540 feet, and probably reaches R. camtschaticum at 6235 feet. We started our October 25th trip with a local climber who left us behind; soon we were into the snow patches visible from our hotel room, but we got well into the R. brachycarpum zone without difficulty. Then heavier, wind-driven snow threatened to obliterate the trail, and our morning companion came down, saying he could not make any progress through deep snow above. Recognizing that we would have little chance of finding the R. camtschaticum , we stopped and contented ourselves with R. brachycarpum . The highest we saw (No. 822, elev. 4625 feet) had no seed, but we found moderate quantities at the following, lat. 39°50', long, 141°01':
No. 823 elev. 4,265 - 4,330 feet
No. 824elev. 3,545 - 3,740 feet
Both the leaf size and the amount of indumentum increased as we descended. No. 822 had only very slight felting near the midrib; No. 823 somewhat more, spreading out in the form of patches on the leaf blade; and No. 824 a still larger amount on some leaves only, forming an almost continuous coating. Other No. 824 leaves of the same maturity were almost as bare as those found higher.) Each of these had considerably less indumentum than any we saw at Mt. Apoi, but No. 824 closely resembled the Sounkyo No. 821, both in leaf size and amount of indumentum.
On the same lwate trip, an azalea at 2375 feet yielded seed - No. 825. This is almost certainly R. kaempferi.
On October 27th we tried 5375-foot Komagatake and possibly reached the zone where this mountain's R. camtschaticum grows - but could not be sure of this because of the snow. The groups of R. brachycarpum with seed (lat. 39°46', long. 140°49') were:
No. 827 elev. 4,380 feet
No. 829 elev. 4,020 - 4,135 feet
Each had relatively small leaves with only slight traces of indumentum along the midrib.
This completed our 1983 work on R. brachycarpum , probably more than meeting all demand for seed. But we were also interested in supplying Dr. Chamberlain with unbiased material from known locations, to help him decide whether to retain his already published intraspecific groups (1982, p. 307) or whether to adopt those we favor. And, if we saw seed, we picked some.

R. makinoi Territory
The day after our Komagatake venture, we traveled via two of the bullet train systems to Toyohaski. Then the next morning, October 29th, we took a local train up the lida Line to Hon-nagashino, the small station that serves as gateway to the R. makinoi country. Arriving at 9:39 as agreed, we were met by Mr. Kenji Matsushita - farmer, retired township official, and prominent member of the Japanese Rhododendron Society's Nagoya Chapter. Mr. Suzuki had asked him to show us as much as possible of R. makinoi's geographic range during our three-day stay, and we were fortunate indeed to have such a mentor.
First spreading out a map on which he had divided into three sections the principal distribution area of this plant, he suggested we try to cover one of these each day, starting immediately. This sounded fine to us, so off we went in his car through the very rough terrain, which looks like dozens of small mountains squeezed together.
Such a squeeze may indeed have occurred, since R. makinoi is evidently a survivor of those rhododendrons present in Japan before the arrival of a large additional piece of land, thought to have moved out from the Asian coast. And this plant's present range is a portion of the extremely complex wedge of mountains which appears to have been compressed by the arrival of the new land. Yet in this small range it grows profusely.

Each of our three search areas was approximately six to seven miles long by three to four miles wide, and we first headed for the east one, which lapped over from Aichi Prefecture (where we were based) into Shizuoka Prefecture. Our first stop was the home of an older gentleman who guided us that day. He began with a nearby trail, leading to the top of a ridge at lat. 34°57', long. 137°39', where at elevations of only 1250-1310 feet we saw numerous plants of R. makinoi in fairly dense shade. Although we looked at many of these, we found no seed and had to be content with vegetative specimens: No. 830. (As with R. brachycarpum , much of our R. makinoi material goes to Dr. Chamberlain to provide him an adequate basis for reappraisal of this taxon. It is with his tentative approval that I treat it as a species rather than a subspecies.)
From this ridge we returned to the older gentleman's home, where Mrs. Doleshy and I learned that he is perhaps the largest commercial supplier of R. makinoi plants in Japan. He had them in all sizes, from seedlings in flats to gnarled, thick-trunked specimens perhaps 200 years old. To propagate his flower color selections, including whites, he was grafting onto rootstocks of the same species. Also he was producing numerous bonsai and specially shaped plants, and he offered to give us one of his best - which we had to decline with thanks. We did, however, look carefully at all his groups of plants, since here if anywhere we might see evidence of variability to be found in the wild. But he was apparently selecting only the variations in flower color and the unusual stamens - prized as horticultural features but of little taxonomic significance in this group of rhododendrons - and evidently had not found anything else that caught his eye. True, the plants respond to sun and shade with differing growth habits and leaves of various sizes, shapes, and degrees of curvature. But these environmental effects are not evidence of any inherited traits, and they may indeed conceal such traits.
From this nursery, we headed east toward the prefectural boundary via a maze of narrow, winding roads. Our next stop was a stand of R. sanctum just over the line in Shizuoka, at an elevation of approximately 900 feet, lat. 34°57', long. 137°42'. We learned that the R. sanctum of this region is found on only one kind of rock, a highly fissured serpentine on which it appeared to thrive. The plants were 8 to 10 feet tall and much-branched, completely dominating a steep slope with few trees. The large leaves were somewhat bronzed at this season (a change we do not see in Seattle), and they filtered the sunlight to an orange-russet color. Our collection was No. 831.
After an excellent lunch at a small, country restaurant, we started looking for R. makinoi in this boundary area. Several brief halts were nonproductive. But we then came to a farm, apparently right on the boundary, and the largest R. makinoi that Mrs. Doleshy and I had ever seen was growing in a hillside hayfield behind the buildings. The lady of the house welcomed us and said her plant had been moved 80 years ago from the hill across the road to its present location. From this plant and from the apparently wild population still found across the road, we collected our No. 832 (elevation 1475 feet, lat. 34°58', long. 137°42'). Since the plant in the hayfield may be close enough to others for exchange of pollen, and since at least two-thirds of our seed came from the plants across the road, we think this seed may provide nearly unbiased representation of the local R. makinoi population. Therefore the resulting plants may be of value for distribution study.
This finished the day's work; the next morning, October 30th, we headed for our western search area. With a local forester as guide, we climbed up the trail to the beautiful Tanayama Plateau. Mr. Suzuki had recommended this as the place to see a large stand of R. makinoi; nevertheless we were amazed to find ourselves among - unquestionably - hundreds of thousands of mature plants. Yet, despite the number of these plants, we did not find a single capsule of current year's seed on the plateau or the surrounding ridges. We did, however, obtain R. pentaphyllum var. nikoense with a small amount of seed as No. 834, Enkianthus perulatus as No. 835, and herbarium specimens of the R. makinoi as No. 833 (all at an elevation of 2300 feet, lat. 35°00', long. 137°35'). Expanding our search for the R. makinoi seed, we took a roundabout trail back to the car but still found none, even though the plants throughout the whole area were healthy and well budded for 1984. Mr. Suzuki, when we talked with him a few days later, found this hard to believe, and we speculated that a sudden chill was responsible.
Driving down, Mr. Matsushita suddenly stopped at an elevation of approximately 700 feet and pointed to a plant beside a huge rock in the middle of the rice fields. This was a R. makinoi , perhaps brought down from the Tanayama Plateau many years ago; it had grown to a diameter of about 14 feet and a height of about 9 feet, our rather short guide being able to walk under it. We took specimens and the relatively few seed capsules as No. 839 (lat. 35°02', long. 137°34'). Since this plant was isolated, the seed is doubtless uncrossed but cannot be attributed to any particular wild population.

We then dropped off the guide, and Mr. Matsushita took us to his attractive house, where Mrs. Matsushita gave us French pastries and coffee. On a side hill around and below the house, overlooking rice and tea fields, is their rhododendron garden. This includes an excellent collection of the Japanese species, with some emphasis on white-flowered forms; with these are many hybrids, some familiar to us and others not. All were growing well.
The next day, October 31st, the destination was our third and last search area, the central one from Mt. Horai east to Horaiko (lake). First we looked near the lake for accessible R. makinoi but found none. Then Mr. Matsushita consulted at the headquarters of the Aichi Prefectural Citizens' Forest. The people there gave us a special permit to drive up a road normally closed to the public, and this led to a splendid but seedless stand at an elevation of 660 feet or less. Therefore we went back to forestry, and this time their suggestion was a trail climbing up a ridge from practically their back door. Until then we had been trying to reconcile ourselves to the seed scarcity. But with the new suggestion from forestry, we received some hint that we might find what we were looking for. As a result, Mr. Matsushita and I climbed from the 295 to the 1017-foot level in about 20 minutes - a climbing rate of more than 2000 feet per hour (which I seldom if ever attained during my years of active climbing). On a large rhododendron at the top of the ridge we at first could see no seed - perhaps being so used to failure that we expected it. Then we saw the first cluster of capsules, and more and more. This is our No. 841 (lat. 35°00', long. 137°38') and represents with no bias whatsoever a wild population. Capsules were small, compared with those on farmers' plants, but they were by no means empty, and we also obtained good herbarium material.

After lunch, Mr. Matsushita drove us up the new Mt. Horai skyline road, which bypasses many of the 1200 steps we had climbed with the religious pilgrims in 1965. We were not overjoyed with the new parking lot for two dozen buses, enabling tourists to check off this attraction without doing any walking at all. But we then drove down to the village at the mountain's foot and found we could still recognize almost everything. Although the little railroad has been replaced by buses and the steps in the single street are paved over, this street has not been widened, and the shops look the same. Also the inn where we stayed looks prosperous (indeed had no rooms available when we tried to make a reservation this year), and the 1200 stone steps still begin just beyond the inn.
When Mr. Matsushita stopped his car suddenly, we looked up and found ourselves under the same street side planting of R. makinoi that had supplied our No. 4 collection of 1965. This year, they were again loaded with seed, which we picked as No. 843 (elevation 600 feet, lat. 34°58', long. 137°35'). And although moved here long ago, these plants almost certainly came from the mountain above; also they are sufficiently close together to exchange pollen, therefore the seed from them probably comes close to giving an unbiased representation of Mt. Horai's wild population.
From there, at the end of our successful final day of collecting, we went to the township offices for photos and interviews. The 72-year-old chief agricultural official turned out to be one of Mr. Matsushita's fellow members in the Nagoya Chapter of the Japanese Rhododendron Society, and with these two we soon left the township headquarters for a more informal rhododendron chat at a nearby coffee shop.
During our three days in R. makinoi territory, we had quite intensively examined an area measuring about 9 7 miles. While additional stands can be found as much as eight miles north and east of those we saw, we are by no means disappointed. Stands at the territorial limits would of course be interesting to visit. But, in our rather broad sample, we cannot see any large amount of variation, and we are not aware of any consistent differences that correlate with locations, therefore we doubt that anything very unusual is to be found in the outlying areas.

End of Journey
Since our rail passes were still good, we took a two-day vacation trip via one of the newer bullet train systems to Echigo-Yuzawa and Niigata - parts of Japan we'd never seen. Then we returned to Tokyo for work on our seeds and an enjoyable day with Mr. Suzuki. He told us that winter had indeed come early to Hokkaido, and he was particularly interested in our searches for R. makinoi . Concerning the difficulty many U.S. growers seem to have with this species, we concluded that they may be unaware of two things: first, that the wild plants grow in practically neutral soil (pH 7), and second, that these come from a climate with peak rainfall during August.
After an excellent lunch at one of Mr. Suzuki's favorite sushi restaurants, we visited the garden of Mr. Oyama, honorary president of the Japanese Rhododendron Society. Although he was hospitalized for an illness, Mrs. Oyama and Mr. Suzuki knew the garden well and showed us the rich assortment of rhododendrons, watched over by ornamental spiders Mr. Oyama had introduced.
The trip back to Seattle, fortunately, was a snooze across four seats each of the 747. This left us in shape to appreciate the kindness and efficiency of the plant quarantine inspector, who cleared our seeds after less than 20 minutes' examination at the airport plant laboratory. Now they have gone to Bill Tietjen for the Seed Exchange.
References
Chamberlain, D.F. 1982. Revision of Rhododendron II, Subgenus Hymenanthes, Notes from Royal Bot. Gard. Edinburgh 39: 209-486.
Doleshy, F.L. 1983. Distribution and Classification of Certain Japanese Rhododendrons. Jour. Am. Rhod. Soc. 37: 81-89.