The systematics of Himalayan willows is very poorly known. Andersson's works (1851, 1860, 1868) still constitute the core of our knowledge about the willows of this region. Data from these works, with some insignificant changes have been the foundation of Salix treatment in Hooker's Flora of the British India (1890); all subsequent authors, in turn, have been relying on Hooker's work. After Andersson and Hooker, there were no significant revisions of Himalayan willows' systematics.

I had to deal with Himalayan willows while clarifying the geography and nomenclature of two species from the USSR flora—S. daphnoides Vill. and S. oxycarpa auct. ross. I also had to treat Salix for Prof. Rechinger's Flora Iranica, which embraced (not quite appropriately) a large part of the region harboring the Himalayan flora.

I've examined Himalayan collections in the Herbarium of the Botanical Institute in Leningrad as well as a fair amount of samples (including a number of types) generously shared by the Royal Botanic Gardens in Kew and Edinburgh, British Museum, National Museum in Stockholm, Center for Botanical Survey of India in Calcutta, National Herbarium in Washington, D.C., and Herbarium of the Florence University. I am deeply grateful to all the institutions and individuals who have provided assistance for this study.

The scrutiny of the named collections has revealed a need for amendments and clarification in the understanding of a number of Himalayan species.

Starting from Andersson, nearly all of the authors have listed S. daphnoides Vill. for the western Himalaya. This is a species described from the French Alps and distributed in Central Europe, the Baltic states, and southern Scandinavia. That was solely Floderus (1935: 311) who found that the Himalayan willow known under the name of S. daphnoides was not identical to the European plant. He described it as a distinct species, S. dolichostachya Flod. However, this point of view was not accepted in later reviews by Hao (1936) and Bor (1953). Whether S. daphnoides is actually distributed in the Himalaya, consequently remains an unresolved question.

The study of the material at my disposal has brought me to the following conclusion: the Himalayan plant is a different species, indeed, not in the least related to the European willow. Even the section to which S. daphnoides belongs is not represented in the Himalaya. Nevertheless, the establishment of a new species by Floderus, was not needed, since "S. daphnoides from the Himalaya" has shown no difference from another species listed in every report from the Himalaya—S. oxycarpa Anderss. However, the name S. oxycarpa, in turn, is to be superseded by a priority name S. sericocarpa Anderss.

M.I. Nasarov (1936: 175) and other authors after him included S. oxycarpa in the USSR flora. However, the comparison of samples from the Soviet Central Asia with the Himalayan material and Andersson's authentics has shown that all of the Central Asiatic samples previously identified as S. oxycarpa actually are S. pycnostachya Anderss., a closely related, though completely distinct species. Identification of S. oxycarpa with S. pycnostachya (Skvortsov 1962: 70; Ikonnikov 1963: 90) has proved to be a mistake.

The following is an amended and enhanced description of S. sericocarpa along with additional comments.

Salix sericocarpa Anderss. 1860, Journ. Linn. Soc. Bot. 4: 43; Hook. f. 1890, Fl. Brit. Ind. 5: 637; C.K. Schn. 1916, in Sarg. Pl. Wils. 3, 1: 112; Görz 1934, Feddes Repert. 35: 286. — S. alba var. eriocarpa Hook. f. et Thoms. in shed. — S. daphnoides var. indica Anderss. 1851, Sv. Vet. Acad. Handl. 1850: 475; ejusd. 1860, l. c.: 46. — S. oxycarpa Anderss. 1860, l. c.: 45; Brandis 1874, Forest Fl. Ind.: 471; ejusd. 1906, Ind. Trees: 638; Hook. f. l. c.: 638; C.K. Schn. l. c.: 171. — S. insignis Anderss. 1860, l. c.: 47, p. p. saltem. — S. daphnoides auct. non Vill.: Brandis 1874, l. c.: 409, tab. 62; ejusd. 1906, l. c.: 637; Hook. f. l. c.: 631; Collett 1902, Fl. Siml.: 480; C.K. Schn. l. c.: 155; Parker 1924, Forest Fl. Punjab: 507; Bor 1953, Man. Ind. Forest Bot.: 109. — S. dolichostachya Flod. 1935, Geogr. Ann. (Stockholm) 17: 311.

[Latin diagnosis omitted]

A tree, sometimes growing rather tall. Previous-year shoots 1.3-2.2 mm thick, glabrous or tomentose in distal parts, tawny or fulvous. Current-year shoots mostly pubescent. Vegetative buds compressed, 2-4 mm long. Floriferous buds drastically different from vegetative, usually positioned at acute angle to shoot, deltoid-ovoid, mostly acute, slightly or scarcely compressed, 4-8 x 3-4 x 2-3 mm, either of same color as shoots or brighter, reddish-tinted. Stipules usually well developed, free (not connate to petioles), up to 10 mm long (on nourishing shoots to 15 mm), obliquely semi-cordate or nearly lanceolate, acute, straight or more or less distorted, dentate at margin. Petioles 6-15 mm long. Blades of well developed leaves lanceolate or elliptic, 60-120 x 10-35 mm. Length-to-width ratio 3-7. Leaves broadest about or just below their middle, cuneate (rarely almost rounded) at base, attenuate at apex, mostly of distinctly contrasting colors on upper and lower surface: green on upper side, more or less glaucous, opaque on underside. Underside surface flat, veins not prominent. Upper surface without stomata or stomata scattered. Leaf margin dentate. Young leaves more or less pubescent, mature usually glabrous. Catkins expand simultaneously with leaves (coetaneous) or somewhat earlier (subcoetaneous); catkin stipes 6-20 mm long, with 2-5 cataphylls or sometimes developed leaves up to 20-30 mm long; staminate 35-50 x 8 mm, pistillate 40-100 (130) x 11-13 mm by the time fruit ripen. Bracts persistent, dark tawny or blackening at apex, obtuse, lanceolate or elliptic, 1.5-2.3 mm long, covered with short, sparse trichomes. Nectary single, truncate, 0.3-0.5 x 0.2-0.3 mm. Stamen filaments partially connate, more or less pubescent at base. Capsule stipitate, lanceolate-conical, glabrous or sericeous; stipe 0.2-1.0 mm, ripe capsule 5-6 mm long. Style 0.5-0.8 mm; stigmas 0.2-0.4 mm long.

T y p u s: "Kashmir, alt. 1,800 m [6000'], 20 IV 1848, Thomson" (pistillate plant identified as "S. alba var. eriocarpa Hook. et Thoms.") (K).

In addition to the type, I examined 34 non-duplicate samples of S. sericocarpa. As labels suggest, this species is rather broadly distributed in Chitral, Kashmir, and Punjab at 1,700-3,500 m, often cultivated.

The type sample (holotype) consists of two branchlets: one with pistillate catkins and young leaves, the other with just young leaves. Attached to the same sheet, there are two more branchlets that belong to completely different species: S. babylonica L. with staminate catkins and S. alba L. s. l. with just young leaves. The confusion apparently resulted in the name S. alba var. eriocarpa applied by Thomson to his specimen. Andersson and later on Schneider placed S. sericocarpa closely to S. alba because of that. Hooker included S. sericocarpa with doubtful species—indeed such an artificial entity could look anything but natural. Nevertheless, since a satisfactory type specimen (lectotype) can be selected, the name S. sericocarpa Anderss. should be preserved, in accordance with the articles 7, 9, and 70 of the International Code of Botanical Nomenclature of 1959.

The plant in question is not at all related to S. alba. On the other hand, one cannot agree with Floderus, who placed it closely to Capreae group. Despite its unusually large leaves, S. sericocarpa should be placed within section Helix Dum., in close proximity to S. tenuijulis Ledeb. and S. pycnostachya Anderss. S. sericocarpa is different from the latter two species in its overall stately habit, broad leaves, mostly nearly semi-cordate stipules, its bud shape reminiscent rather of S. caprea buds, and larger catkins. Leaves of S. sericocarpa are more conspicuously bicolored than in S. tenuijulis or S. pycnostachya; stomata on the upper leaf surface are either lacking or sparse (rarely more dense), as opposed to the other two species, whose leaves are always densely dotted with stomata. Margin dentation is also different. Stamen filaments in all available staminate samples of S. sericocarpa were only partially connate, whereas in S. pycnostachya and S. tenuijulis stamen filaments are always completely connate. Partially connate stamens served as diagnostic character for Hooker and Thomson when they identified a willow sample as S. zygostemon Boiss. The latter species described from Iran is still very poorly known due to its extreme rarity. However, we can confidently state that S. zygostemon is not S. sericocarpa. S. sericocarpa differs from species of sect. Daphnoides in many ways. For example, it does not have bluish boom on the branches (except sometimes upon drying); stipules are free (and not connate to petioles as in sect. Daphnoides); leaf-margin dentation is of a different character; catkins are stipitate, not sessile; bracteal pubescence is not much pronounced, whereas ovaries are rather pubescent; stamens are partially connate and more or less pubescent, their anthers rather small. This list of differences is not exhaustive.

We don't know much about S. insignis Anderss. The description of this species might have been also based on mixed material. Leafy samples supposedly were those of S. sericocarpa (this is confirmed by characters listed in Andersson's description as well as some of his identifications on herbarium samples.) However, the flowering samples might belong to other species. Samples received from Kew as authentics of S. insignis contain just pistillate and staminate catkins without leaves. The staminate catkins, which do match Andersson's description, probably belong to S. aegyptiaca. The pistillate catkins, which are completely inconsistent with the description, apparently belong to S. turanica Nas.

S. hastata L. appears listed in all reports from the western Himalaya starting from the very first Andersson's work (1851: 479), sometimes under the name S. hastata L. var. himalayensis Anderss. Floderus (1935: 306) did treat this plant as a separate species, S. himalayensis Flod. However, as it happened to "S. daphnoides from the Himalaya," Floderus' opinion was not taken into account in later works by Hao (1936) and Bor (1953).

I have examined more than 70 non-duplicate samples of "S. hastata from the Himalaya," including an isotype—a sample from Royle's Herbarium preserved in Leningrad under the name "S. hirta Royle." Upon studying this large material, I was able to conclude that Floderus had been quite right when making a difference between the Himalayan plant and Euro-Siberian S. hastata. Differentiating characters were correctly marked by Floderus, so there is no need to repeat them here. Having said all that, we still have to synonymize S. himalayensis with S. karelinii Turcz., due to their complete similarity.

Together with S. himalayensis, Floderus (1935: 310) published another new species from the Himalaya: S. adenophylloides Flod. He described this species while basing solely on a single sample, whose photo was enclosed with the description. Upon examining the photograph and reading the description, I can tell with confidence that the plant in question is just a specimen of S. karelinii with unusually long (to 12-14 cm) pistillate catkins.

Consequently, the synonymy is going to look this way:

Salix karelinii Turcz. ex Stschegl. 1854, Bull. Soc. Nat. Moscou 27: 196; Turcz. op. cit.: 393; A. Skv. 1962, Bot. mat. Gerb. Inst. bot. AN UzSSR 17: 63. — S. prunifolia Kar. et Kir. 1842, Bull. Soc. Nat. Moscou 15: 452, non Smith 1804. — S. hastata auct. non L.: Anderss. 1851, Sv. Vet. Acad. Handl. 1850: 479; ejusd. 1860, Journ. Linn. Soc. Bot. 4: 51; Brandis 1874, Forest Fl. Ind.: 467; Hook. f. 1890, Fl. Brit. Ind. 5: 630; Collett 1902, Fl. Siml.: 480; Parker 1924, Forest Fl. Punjab: 506; Hao 1936, Syn. Chin. Salix: 83; Nas. 1936, Fl. SSSR 5: 116, p. p. quoad Pl. As. Med.; Bor 1953, Man. Ind. Forest Bot.: 109. — S. hastata var. himalayensis Anderss. 1867, Mon. Salix: 173. — S. himalayensis Flod. 1935, Geogr. Ann. (Stockholm) 17: 306. — S. adenophylloides Flod. l. c.: 310.

S. karelinii is encountered quite often in the subalpine and alpine zones on most mountain ridges of the Tien Shan, Pamir-Alay, and Himalaya: in Nuristan, Chitral, Kashmir, the Himalayan part of Punjab and Union Territories, and in northeastern Nepal. Within the USSR territory, it ascends to 3,600 m, while in the Himalaya it reaches 4,500 m.

Among other newly described species published by Andersson in 1860, there were S. fruticulosa Anderss. and S. serpyllum Anderss. The author placed the two in different sections. It is not quite clear why he later (Andersson 1868) renamed the former species as S. furcata Anderss. Hooker (1890: 634) considered S. serpyllum valid, while treating S. fruticulosa as a vague, doubtful species (l. c.: 637). However, Schneider (1916: 119, 147) rehabilitated S. fruticulosa Anderss. He considered both species as valid and again placed them in different sections. Indeed he circumscribed a separate new section solely for S. serpyllum.

I have examined the type material for both S. fruticulosa (Strachey et Winterbottom, # 13, LE) and S. serpyllum (Sikkim, Lachen, 14,000', 15 VII 1849, Hook. f., LE, BM, E, K, S) as well as a large amount of subsequent collections (mostly received from the British Museum as unidentified samples) and verified with complete certainty the identity of the two species to each other. The description of the two different species could be explained by the fact that Andersson was dealing with samples that had been growing in different conditions. The one that grew in a more favorable situation (S. fruticulosa type) is a plant with developed ascending stems, which superficially somewhat resembles S. coesia Vill. The plant sampled in a less favorable environmental situation (S. serpyllum type) produced stems appressed to the ground, all covered with adventitious roots, with only some young shoots ascending. Additionally, the samples were at different stages of development.

The nomenclature, therefore, is going to look as follows:

Salix fruticulosa Anderss. 1860, Journ. Linn. Soc. Bot. 4: 53; C.K. Schn. 1916, in Sarg. Pl. Wils. 3, 1: 119. — S. serpyllum Anderss. 1860, l. c.: 55; ejusd. 1868, in DC. Prodr. 16, 2: 292; Hook. f. 1890, Fl. Brit. Ind. 5: 634; C.K. Schn. l. c.: 147. — S. furcata Anderss. 1868, l. c.: 291; Parker 1924, Forest Fl. Punjab: 508.

T y p u s: "Kumaon, Pindari, 12,000', Strachey et Winterbottom, # 13," [staminate and pistillate] (K, isotypus LE).

The species is distributed in the eastern Himalaya (Kumaon, Nepal, Sikkim, Bhutan) at elevations from 2,700 to 4,500 m, mostly at open moist rocky or grassy habitats, on screes, moraines, often in low rhododendron thickets.

There is yet another Andersson's species from the Himalaya that has not been clarified. This is S. myricaefolia Anderss., which was later renamed for an unknown reason by the author himself as S. divergens. Andersson noticed the similarity of this species with the European S. coesia Vill.; however, in all subsequent literature the Himalayan plant has been treated as a distinct species. On the other hand, up until now the West European authors have been restricting the area of S. coesia exclusively to Europe. However, in the Russian literature, starting from Trautvetter (1832: 305), S. coesia has been listed for south Siberian mountains and Central Asia as far as the Pamirs.

Comparison of Himalayan samples of S. myricaefolia with our samples from Central Asia and European S. coesia has proved completely impossible treating any of these as different species. The synonymy, therefore, should look as follows:

Salix coesia Vill. 1789, Hist. Pl. Dauph. 3: 768; Trautv. 1832, Nouv. Mem. Soc. Nat. Moscou 2: 305; Kar. et Kir. 1842, Bull. Soc. Nat. Moscou 15: 452; Turcz. 1854, op. cit. 27: 394; Nas. 1936, Fl. SSSR 5: 177; A. Skv. 1962, Bot. mat. Gerb. Inst. bot. AN UzSSR 17: 68. — S. myricaefolia Anderss. 1851, Sv. Vet. Akad. Handl. 1850: 483; ejusd. 1860, Journ. Linn. Soc. Bot. 4: 53; S.K. Schn. 1916, in Sarg. Pl. Wils. 3, 1: 172; Parker 1924, Forest Fl. Punjab: 510; Hao 1936, Syn. Chin. Salix: 109. — S. divergens Anderss. 1868, in DC. Prodr. 16, 2: 316; Hook. f. 1890, Fl. Brit. Ind. 5: 637. — S. minutiflora Turcz. ex E. Wolf 1903, Acta Horti Petropol. 21: 141; Kryl. 1930, Fl. Zap. Sib. 4: 737. — S. sibirica Ledeb. 1850, Fl. Ross. 3, 2: 622, p. max. p. saltem, haud Pall. 1788.

T y p u s: "Dauphiné," Villars (verosimiliter in Herb. Grenoble, isotypus LE).

Within the Himalaya, this species is distributed only in the arid extreme northeast adjacent to the Pamirs.

 

Andersson N.J. 1851. Ost-Indiens hittills kända pilarter. Svensk Vet. Akad. Handl. 1850.

Andersson N.J. 1860. On East Indian Salices. Journ. Linn. Soc. London (Bot.) 4.

Andersson N.J. 1868. Salix. In A.P. De Candolle. Prodromus systematis naturalis regni vegetabilis 16 (2). Parisiis.

Bor N.L. 1953. Manual of Indian Forest Botany. Bombay.

Floderus B. 1935. Some new Himalayan willows. Geogr. Ann. 17. Stockholm.

Hao Kin-shen 1936. Synopsis of Chinese Salix. Feddes Repert. Spec. Nov. Regni Veg. Beih. 93.

Hermann F. 1956. Flora von Nord- und Mitteleuropa. Stuttgart.

Hooker J.D. 1890. Flora of the British India 5. London.

Ikonnikov S.S. 1963. Opredelitel rasteniy Pamira [Guide to the Plants of the Pamirs.] Dushanbe.

Nasarov M.I. 1936. Salix. In Flora SSSR [Flora of the USSR] 5. Moscow—Leningrad.

Rechinger K.H. 1957. Salix. In G. Hegi. Illustrierte Flora von Mitteleuropa 3 (1). München.

Schneider C.K. 1916. Salicaceae. In C.S. Sargent. Plantae Wilsonianae 3. Cambridge.

Skvortsov A.K. 1962. [Synopsis of willows from the Soviet Central Asia.] Bot. mat. Gerb. Inst. bot. AN UzSSR 17.

Trautvetter E.R. 1832. De Salicibus frigidis. Nouv. Mém. Soc. Nat. Moscou 2.

---

Translation I.Kadis
23 Aug 2007