A Northumberland Alpine Gardener's Diary
This entry: 27 August 2011 by John Richards
Chinese Primulas. Entry 191.
Regular readers will know that I have used the last few issues to cover a number of groups of plants that we saw during our visit to China in late June and early July, and may be surprised that I have not yet mentioned any primulas. Well, yes, we did of course see many different primulas, about 34 species I think, and one of the reasons that I have not mentioned them before is that I am still editing pictures taken during the trip and mulling over problems that a number of the primulas we saw posed.
I shall not show a number of species which were often common and spectacular, but did not raise any questions: P. sikkimensis and P. secundiflora for instance, P. involucrata, P. polyneura, P. septemfida, P. cockburniana, P. deflexa, or species that had finished flowering; P. sonchifolia in particular. However, some non-problematic plants were so stunning that I can't resist picturing a few. I had seen the wonderful high alpine, P. dryadifolia, a good candidate for the best primula of all, on two previous occasions, but always past its best. We were fortunate to see it in good form on both the summit of the Hong Shan, and on the eastern side of the Da Zue Shan. Here is a particularly good plant at the former site.
Another wonderful plant that I had seen before became so much part of our high alpine experience, both in Sichuan and later in Yunnan that we were in danger of treating it with overfamiliarity. However, P. nanobella is hard to resist! Notice by the way that I am presently treating all plants from the north and east of the range of the bella aggregate, including all those seen on the current tour, as P. nanobella, distinguished by usually (but not always) having a lilac furry eye, rather than a white one, and short flower stalks. The true P. bella is much more like the Himalayan P. primulina and occurs further south and east, on the Cang Shan (and Yulong Shan?) and westwards into Myanmar and Tibet.
A plant that had previously caused controversy, has now settled down as P. boreiocalliantha, although it had been named fairly recently as a new species, P. hongshanensis. The plant we thought was P. boreiocalliantha is in fact correctly named P. bryophila! It is shorter and pinker in colour than the present stately subject which is common at over 4000 m on both the Hong Shan and the Da Xue Shan.
A factor that links all the previous subjects is their virtual intractability in cultivation, and the same goes for the next plant featured. P. amethystina is commonly encountered in wet peaty ground with flowing water at over 4000 m almost anywhere in western China. Normally it occurs as its handsome variety brevifolia. The type variety is apparently confined to the type locality for the species, the Cang Shan above Dali. Here is v. brevifolia growing near the Huai Hai Zi lake south of Kangding, although we saw it in many places.
However, on the Zi Chou Pass above Jiulong, and later on the Hong Shan, 'normal' plants of P. amethystina were accompanied by much more slender individuals with smaller narrow bells and small round leaves with small marginal teeth. These answer the description of v. argutidens, but in my view they form part of the variation of v. brevifolia. More intriguingly, I can see no difference between these slender individuals and the little-known species P. silaensis, which was originally described from plants collected in NW Yunnan, on the Myanmar border. Such plants have been photographed there recently, and in my view they are the same as P. amethystina v. argutidens. I think we can lose P. silaensis!
Lets move on to the Muscarioides, of which we saw no fewer than six species. I start with P. muscarioides itself, which takes a fairly central position in the section. We saw this on several occasions both in Sichuan and Yunnan. Here it is in a handsome form on the Hong Shan.
Closely related, but a taller plant with lobed leaves and more widely open flowers is the stately P. cernua. We saw this growing together with P. muscariodes on the Zi Chou Pass, north of Jiulong. A similar plant grows with P. longipetiolata on the Zheduo Pass near Kangding, but the latter has glandular stems and may be an undescribed species.
We has been told by Dieter Zschummel and Pam Eveleigh that a mysterious plant grew with P. muscarioides towards the north-eastern end of the high pass over the Hong Shan. Using the wonders of GPS, we found these populations, 'in the pink'. After a close examination of the photos I have come to the opinion that P. apoclita grows with P. muscarioides here, but there are some intermediate plants so that the two species may well hybridise here.
Here are some of the intermediates.
We move on to P. florida, which is a very common plant indeed, growing in dense colonies on rocks and banks in the forest zone on all the mountains we visited. Usually it has quite dissected leaves with short petioles, but we found two variants. Here first is a typical plant.
By the Zi Chou Pass we found distinctive populations growing in more of a tundra-type habitat at nearly 4400 m. These were very dwarf and had almost entire leaves. Possibly they deserve naming as a race of P. florida.
On another high pass, near the Huai Hai Zi lake south of Kangding, were plants with longer petioles and skeletally dissected leaves. These answer the description of P. laciniata, but I think they can be safely encompassed within the variation of P. florida, which we found to be very considerable. Another species bites the dust!
poissonii and wilsonii
On the Zi Chou Pass, north of Jiulong in west-central Sichuan we came across plenty of a 'candelabra' primula (section Proliferae) that at the time we called P. poissonii. However, that familiar species is chiefly Yunnanese in its distribution, and when we did find it east of Zhongdian (it is common on the Zhongdian plateau and south to Lijiang) I realised that our Sichuan plant had in fact been P. wilsonii. P. wilsonii is variable; in some forms, previously called P.anisodora, the leaves are unpleasantly aromatic and the flowers often almost black. Others approach P. poissonii much more closely, but have a very short flower stalk, usually shorter than the calyx, and the leaves are much bigger, bullate, toothed and ovate-acute. P. poissonii has rather blueish smooth oblanceolate leaves with crenate teeth and longer pedicels. Here, first, is a picture of P. wilsonii just north of Jiulong, followed by P. poissonii near Zhongdian.
In this account, we have already 'disposed of' two species, P. laciniata and P. silaensis. Now comes the turn of another little known and insecure taxon. P. fernaldiana is reputed to be a relative of P. pulchella (section Pulchella) from northern Sichuan. During our 2007 trip I found a single plant which fitted the description of this enigmatic species well, incongruously growing in a wet dtich by a main road near Danba close to a roadside cafe. Basically it is considered to differ from its relatives by long pedicels to the large flowers which thus become pendant, and by the relatively broad, obtuse leaves.
However, when we discovered a large population of primulas growing beside the road north of Jiulong on a wet cliff, it was immediately evident that the status of P. fernaldiana was open to question. Most of the plants were small with narrow leaves and short pedicels and were clearly the common P. pulchelloides which is commonly found in such places, especially in southern Sichuan and northern Yunnan. However, gross, well-developed individuals matched the description of P. fernaldiana well, but were clearly part of the same population. Another one bites the dust! Here is a picture of part of the population, followed by 'fernandiana types'.
While we are about it, here are two more species allied to section Pulchella which are not well-known. First, a plant we only found growing with the rash of orchids on the road from Jiulong to the Wuxu lake (entry 189). Here were a few plants of the handsome P. conspersa. Like most of the relatives of P. gemmifera, this seems to have been little-known to our predecessors, but is a distinctive and interesting plant which has the potential to be persistent in cultivation because the base breaks up into small resting buds which perennate. The closest relative is P. zambalensis, best known from the Bai Ma (Beima) Shan which has bluer flowers and is rarely 'candelabroid', but superimposed whorls of flowers seems to be the norm for P. conspersa. It has entered cultivation briefly under the soubriquet 'SSSE20' in a form with distinctively black smurdged leaves. but it does otherwise seem to be the same thing.
P. stenocalyx is a quite different thing, a small chasmophyte which has the air of a classy European Auricula, P. glutinosa perhaps. I first found this plant on rock faces near the summit of the Zheduo Pass back in 2007 when it gave me several headaches and I thought it might well be new. At the time I considered its closest relative to be P. stenocalyx from northern Sichuan, but that species seemed out of range and ecologically distinct, a subalpine from wet cliffy springs. I was inclined to describe the Zheduo plant as a new species, but two correspondents whose opinions I respect both thought it might be P. stenocalyx, and after a second visit and a much closer examination I am inclined to agree, although it has narrower leaves, unicoloured sepals, and is at best a very distinct high alpine ecotype.
We now pass, finally, to the great beautiful morass of the 'nivalids' (correctly section Crystallophlomis). We found two yellow-flowered species with reflexed petals which are therefore placed in subsection Maximowiczii. We were lucky to find the more familiar of these, P. szechuanica, in a wet copse beside the road as we walked down the northern side of the Hong Shan. This is a bigger plant that the next-named, but is exceptionally variable, as the two sets of leaves figured second and third here show.
In this context, it is interesting to compare a plant that we found much further north, by the Huai Hai Zi Lake south of Kangding, also buried in the depth of a wet copse. This was much smaller, and in some cases the shorter, never candelabroid, inflorescences scarcely exceeded the rounded, often glaucous (below) leaves. At the time I called one of these latter populations P. handeliana and the other P. szechuanica. I am now convinced that both the Huai Hai Zi populations were the same, but are they also the same as the much bigger Hong Shan plants? The jury is out, but I doubt if P.handeliana deserves much more than subspecific rank.
Moving on to the P. chionantha aggregate, I was pleased to see, particularly on the Hong Shan, a wide variety of forms which linked the robust, wide-leaved subsp. sinopurpurea (P. sinopurpurea) with the dwarfer, narrow-leaved subsp. sinoplantaginea (P. sinoplantaginea). On previous visits I had seen too few convincing examples of the latter to be convinced by my own convictions (mostly derived from herbarium studies) that the two are merely extremes in a very wide range of forms. What is common to all these types are the strict, erect sheaves of leaves and the sheathing leaf bases, as well as a distinctive calyx and rounded purple corolla lobes ('petals'). The following pictures, all from the Hong Shan at high elevation, show firstly sinopurpurea, then typical sinoplantaginea, and finally an intermediate. Mostly, these plants had finished flower or nearly so during our visit, but this did not interfere with our judgement.
What is Primula diantha?
In this account, I have already suggested that three little-known species of primula, P. laciniata, P. silaensis and P. fernaldiana can probably be disposed of. The same can probably be said of a fourth species, the tiny nivalid P. diantha. P. diantha, originally described from a remote region of western Sichuan, has always been suspect as there is little to distinguish it apart from its very small size and few (one to two) flowers, so the suspicion that it merely represented an extremely reduced relative of P. chionantha has been around for some time.
So it was with some pleasure that a lunch stop towards the south-western end of the long high road which represents the Hong Shan Pass (only negotiable in four-wheel drive vehicles, and well-worth the trouble) revealed not only abundant sinoplantaginea types, but also, growing nearby, tiny one- to two-flowered plants that were indistinguishable from the description and herbarium specimens of P. diantha. Although the type locality of P. diantha, near Batang, needs to be visited, it is hard to avoid the conclusion that P. diantha can be safely sunk within P. chionantha subsp. sinoplantaginea.
The following picture represents an intermediate between 'diantha'and more typical sinoplantaginea.
Discussion of the narrow-leaved relatives of P. chionantha raises the question of P. graminifolia. This name had been lost as a synonym of P. chionantha subsp. sinoplantaginea until recently resurrected by Czech seed collectors. It has proved relatively easy to raise to flowering (although not to keep!), and I have flowered it during both the last two springs (see entry for March 21st this year). As a young plant it has linear leaves, but at flowering the following spring the leaves were distinctly wider and obtuse at the tip, producing a plant that was nearly indistinguishable from P. minor. This agrees well with herbarium material labelled P. graminifolia.
We found the same plant on both the Hong Shan and Da Xue Shan during the recent trip, and this reinforced my opinion that P. graminifolia is related to P. minor, not to P. chionantha. It forms much bigger clumps of leaves than the P. chionatha aggregate, and these are floppy, not stiff, so that the leaf apices typically recurve, which is never seen in the erect leaves of P. chionantha agg.. Also, the habitat is different. P. minor is a plant of wet, shady cliffsand rocks, often near waterfalls, and usually on limestone, and we only found P. 'graminifolia' in steep wet sites, usually flushed road cutting edges. In contrast, the P. chionantha group usually occur in the open, on level or sloping sites. Probably the best solution is for P. 'graminifolia' to be recombined as a narrow-leaved variant of P. minor, v. graminifolia.
P. russeola resurrected?
One of the species of Asiatic primula I discussed during my lecture to the International Conference in Nottingham in April (published in 'The Alpine Gardener' 79: 76) was P. russeola. Most plants so-named in recent years (including those collected by the ACE expedition of 1994 from the Bai Ma Shan) proved to be P. minor and I have experienced difficulty in finding convincing images of true P. russeola, or even in deciding what its distinctive features are.
The first time we encountered a nivalid primula on the current trip was beside a stream high on the Zi Chou Pass, nortb of Jiulong in west-central Sichuan. It was over flower, and after a heavy rainstorm the previous day was in scruffy condition. At the time I named it P.chionantha subsp. sinopurpurea, but since I have studied the photos closely, I wonder? One of the features of P. russeola is said to be the very long calyx (to 15 mm) dissected to two-thirds, and the present plant certainly has long calyces with deep narrow divisions. Also, the inflorescence is covered with rusty glands which is not typical of P. chionantha. This feature is not mentioned in the description of P. russeola either, but then, why else would it be called 'russeola'?. A weak argument I know! The stiff narrow silvery leaves do not match what we are led to expect of P. russeola, and indeed best match those of P. melanops, although the flowers are more like sinopurpurea! No-one said nivalids are easy (although all are beautiful) and I think the identity of the Zi Chou plant will remain unsolved for the present at least.