HARVARD PAPERS IN BOTANY 3(1): 109-112. 1998.

Taxonomic notes on Betula - II.

Two Chinese birches described by A. Franchet in 1899 and misunderstood by subsequent authors

Alexei K. Skvortsov

Main Botanic Garden, Russian Academy of Sciences, 127276 Moscow, Russia.

Abstract. The taxonomic and nomenclatural fate of Betula delavayi and B. fargesii have been unclear since these two species were first described in 1899. The differences between the two and their relationship to other Chinese birches have been treated variously by subsequent authors but never distinctly enough, nor with confidence. Specimens of Betula have been examined at A, GH, LE, MO, NY, and W, and I am confident that these species should be delimited. Betula delavayi and B. fargesii are not the closest relatives of each other; the closest relatives of the former are B. chinensis and B. globispica; those of the latter are B. potaninii, B. calcicola, and B. chichibuensis.


Adrien R. Franchet described in 1899 Betula delavayi from Yunnan and B. fargesii from Sichuan. Besides the types in Paris, no other specimens were cited in the protologue, and apparently no duplicates were available for distribution to other herbaria. The next treatments of Chinese birches by Burkill (1899) and Winkler (1904) were based on inspection of the very same specimens preserved in P. The only addition made by Burkill was the citation of a specimen, Henry 6879, from W. Hubei, which he referred to as B. fargesii (in fact, this specimen belongs to B. delavayi). Regarding B. delavayi, Burkill (1899: 498) commented that "it is connected with B. utilis Don through the var. pratii of the latter. Probably B. potaninii is very closely allied to this species." These statements are also entirely erroneous and were the cornerstones for all subsequent confusion.

Somewhat later, W. W. Smith (1915) described two varieties under B. delavayi: var. forrestii and var. calcicola. Apparently, Smith did not see Delavay's original specimen; still he supposed it should be something intermediate between var. forrestii and var. calcicola. Schneider (1912: 884) made B. delavayi a variety of B. chinensis Maxim., but later (1916: 460) treated it as a distinct species "somewhat intermediate between B. potaninii and B. chinensis, being probably most nearly related to the former." Handel-Mazzetti (1929: 20) raised Smith's var. forrestii to species rank as B. forrestii (W. W. Smith) Hand.-Mazz. and referred var. calcicola to this species. Hu (1948) followed Handel-Mazzetti's treatment. Finally, Li and Cheng (1979) relegated B. fargesii to varietal status within B. chinensis, and B. forrestii to synonymy under B. delavayi; var. calcicola was elevated by Li to species rank as B. calcicola (W. W Smith) P. C. Li.

Thus, the taxonomic and nomenclatural history of Franchet's two species appears to be a perplexing tangle, involving six names at specific rank: besides B. delavayi and B. fargesii, there are B. chinensis, B. potaninii, B. forrestii, and B. calcicola.

This paper attempts to clarify the confusion. It is based mainly on a study of materials preserved in the Harvard University Herbaria (A, GH). I have also had the opportunity to examine the collections of LE, MO, NY, and W; however, their holdings pertaining to the present problem are very limited. Photographs of the types of B. delavayi and B. forrestii were kindly provided by P and E, respectively.

The first and main task of the study was to sort out the relevant specimens from the bulk of all materials of eastern Asian birches, irrespective of the names the sheets bore. Accordingly, I begin my narrative with an enumeration of detected and examined specimens.

Betula delavayi Franch., J. Bot. (Morot) 13(7): 205. 1899. TYPE: CHINA. "China occidentalis; Yunnan in silvis ad Koutoui supra Mo so yn, alt. 2800 m" (HOLOTYPE: P; photographs: A, MHA!). On the label of the type specimen, a somewhat more elaborate handwritten text is legible: "No 3725. Betulus [sic], arbrisseau. Les bois de Kou-toui, au dessus de Mo-so-yn, a 2800 m d'alt. Le 24 Mai 1889 legi ipse ... Delavay."

Additional specimens examined: CHINA. Sichuan: Mount Omei, 2000 m, F. T. Wang 23667 (A); W. Sichuan, W of Kuan-hsien [Guan Xian], Pan-lan-shan, 7--11,500 ft, VI. 1908, E. H. Wilson 3365 (A, MO); W Sichuan, Mupin, 8--9500 ft, IX 1910, E. H. Wilson 4382 (A); E. Sichuan, distr. Tchen-keou-tin [Kangding], R. P. Farges 1011 bis (A); W Hubei, Fang Hsien, A. Henry 6879 (A); W Hubei, Wan Tsao Shan, 7000 ft, W. Y. Chun 3906 (A), same locality, 6200 ft, W. Y. Chun 3957 (A); W Hubei, Shennongjia Forest district, 2100 m, 1980 Sino-Amer. Exped. 183 (A), same locality, 2300--2600 m, 1980 Sino-Amer. Exped. 904 (A, NY).

Of the 10 collections I have seen, half of them are from Hubei. Strikingly few collections were made by the most successful and prolific collectors: for example, E. H. Wilson made two collections; T. T. Yü made one; and J. F. Rock, G. Forrest, and C. K. Schneider made not a single one. Of all the specimens, only three (E. H. Wilson 4382, W. Y. Chun 3906, 3957) had the correct determination, "B. delavayi," probably given by the collectors themselves.

Although the available materials are not very extensive, they still allow a rather definite characterization of the species to be compiled as follows:

Betula delavayi is a small or medium-sized, rather slender tree to 12 m tall. The bark is (dark) gray, rough and sometimes longitudinally fissured. The shoots are usually pubescent with longitudinally adpressed hairs when young but are later glabrous. The petioles are 5--15 mm long. The leaf blades range from 4.2 to 7.2 cm long and from 1.7 to 4.2 cm wide. They are ovate to lanceolate with acute apices, rather uneven marginal serrations, and 8--11 pairs of lateral nerves; the lower sides of the leaves have minute, often early-disappearing glands; the petioles and veins bear longitudinally stretched hairs. The fruiting aments, on very short peduncles that bear 1 or 2 (or none) small leaves, are oval in general outline at maturity and vary in size to 27 x 20 mm. The bracts are up to 12 mm long with lateral lobes similar to the median lobe, narrow, and with rather long, lax pubescence (Fig. 1A--B). The nutlets are more distinct and stable and in general circumscription are mostly broadly oval, 2--3.5 mm long and 1.8--2.3 mm wide; the sides are flattened and form leathery rims 0.1--0.3 mm wide, which taper and disappear toward the base and apex of the nutlet (Fig. 2A--B).

Here it is important to stress that these leathery, nontransparent rims, although usually referred to as "wings," are structurally very different from what we call wings in the white-barked birches. The wings of B. utilis, B. albo-sinensis, B. alnoides, and the American B. papyrifera, for example, have variable and uneven outlines and are very thin, semitransparent, and mostly wrinkled.

Toward winter, the bracts of B. delavayi are shed and the fruiting ament decomposes.

Betula fargesii Franch., J. Bot. (Morot) 13(7): 205. 1899. TYPE: CHINA. "China occidentalis, provincia Sutchuen, ad Heoupin in vicinitate Tchen keou tin, alt. 2200 m (R. P. Farges, N 1012)" (HOLOTYPE: P; photograph and fragment of the type: A!).

Additional specimens examined: CHINA. West China: June 1904, E. H. Wilson 4491 (A). Sichuan: Muli, Mt. Mitzuga, J. F. Rock 24049 (NY, MO), J. F. Rock 24516 (NY); Muli, Wachin, T. T. Yü 14404 (A), 14722 (A); S Sichuan, Yenyuan Hsien, inter Kalapa et Liuki, C. K. Schneider 1198 (A); S Sichuan, mons Taching prope Tu yung pu, C. K. Schneider 4127 (A); Sikang, Djer mai, Tsa-warung, C. W. Wang 65728 (A, LE). Yunnan: E flank of Lichiang [Lijiang] Range, G. Forrest 5546 (type of B. delavayi var. forrestii W. W. Smith), photograph (A, MHA). SW of Lichiang: G. Forrest 21202 (A); Likiang [Lijiang] Range, G. Forrest 22230 (A); Likiang Snow Range, R. C. Ching 3017 (A), T. T. Yü 15020 (A); prope Likiang, C. K. Schneider 1901 (A), 3418 (A), H. Handel-Mazzetti 4165 (W, A, NY); W slopes of Likiang Range, J. F. Rock 3804 (NY), 4173 (A), 8192 (A), 9460 (A, NY); S of Likiang, Sungkwe Hochin Range, J. F. Rock 8318 (A, NY); between Likiang and Youngning, J. F. Rock 17208 (A); Likiang Hsien, C. W. Wang 70988 (A, LE); Chien-chuan-Mekong divide, G. Forrest 22036 (A); Ma-on-shan, K. M. Feng 315 (A); N flank of Haba Snow Range, K. M. Feng 1260 (A); Yunnan, without further information, T. T. Yü 5611, 5630, 6935, 11571 (all A). Xizang ("E Tibet"), SW China (unspecified locality), G. Forrest 23502 (W)).

The herbarium label of the type specimen gives the date of collection as 21 May 1892. The fragment of the type, together with the photograph, were brought by A. Rehder in 1930 from Paris and consist of a twig with three leaves and half of a fruiting catkin.

Of the 30 collections, represented by 37 sheets, not a single one of them had been identified as B. fargesii. About half had been identified as B. forrestii and some others as B. delavayi; about a third had no determination at all. Because the type of var. forrestii W. W. Smith (G. Forrest 5546) belongs to B. fargesii, B. forrestii (W. W. Smith) Hand.-Mazz. should be recognized as a simple synonym of B. fargesii, not as a synonym of B. delavayi as proposed by Li and Cheng (1979:133).

In spite of the rather imposing number of collections, the distribution area of B. fargesii seems to be quite restricted---limited to SW Sichuan and NW Yunnan, that is, to the southern part of the Hengduan mountain region. According to the collectors' notes, the span of altitudinal distribution is from about 2500 to 3800 m. The plants are most often small shrubs and more rarely trees (up to 12 m tall), with reddish-brown trunks. In size and in pubescence of the shoots, buds, and leaves, B. fargesii looks rather similar to B. delavayi, but the shape of the leaves is more ovate and less acute, sometimes even oval with a blunt tip, and the marginal teeth are shorter and more even (see Fig. 3, A--B for B. delavayi versus C--D for B. fargesii). The 8--13 pairs of lateral veins are impressed even more strongly than in B. delavayi. The bracts of the fruiting catkins become strongly lignified toward winter and are not shed (Fig. 1C--D). The ament overwinters and gradually decomposes over the next 1--2 years. The nutlets have very narrow (0.05--0.15 mm wide) but quite distinct leathery rims, usually of the same color as the nutlet itself---brown or reddish brown. The nutlet is usually orbicular in general outline and measures (including the rims) 2.0--3.2 mm x 1.5--2.5 mm (Fig. 2C).

A certain similarity between B. delavayi and B. fargesii does indeed exist, I believe, and both species may be referred to section Chinensis. Still, the materials I have examined do not bear witness to any transgression between the two species. I also see no good reason to distinguish any varieties within these species. On the other hand, both B. delavayi and B. fargesii show very clear affinities to other eastern Asian species.

The nearest relatives of B. delavayi are B. chinensis Maxim. in northern China and B. globispica Shirai in Japan. The nearest relatives of B. fargesii are B. potaninii Batalin and B. calcicola (W. W Smith) P. C. Li in the same Hengduan Mountains and B. chichibuensis Hara in Japan.

Betula potaninii and B. calcicola seem to intergrade, at least to some extent. Their distribution areas appear vicarious; B. potaninii is mainly in western Sichuan and in part of Gansu, and B. calcicola is in Yunnan and southernmost Sichuan. Still, I believe their distinction at the species rank is acceptable. Betula fargesii is not vicarious to any of them and is apparently well isolated.


I am most obliged to Dr. Peter H. Raven, director of the Missouri Botanical Garden, for the support that has made preparation of this paper possible. My sincere thanks are due the herbaria A, E, GH, LE, MO, NY, P, and W for supplying materials for my studies and manifold friendly assistance. Special gratitude is due Mrs. Raisa Trokhinskaya and Dr. Michael Ignatov for typing the manuscript, Mr. Boris Sosnovski for preparing the photograph, and Dr. Anthony R. Brach for various forms of assistance in the course of my work.

Literature cited

Burkill, J. H. 1899. Cupuliferae: Betuleae. J. Linn. Soc. Bot. 26: 496--500.

Franchet, A. 1899. Plantarum sinensium ecloge tertia. J. Bot. (Morot). 13: 197--208.

Handel-Mazzetti, H. 1929. Symbolae Sinicae. Vol. 7, 1. J. Springer, Wien.

Hu, H. H. 1948. The Silva of China. Vol. 2, 1. China.

Li, P. C., and S. H. Cheng. 1979. Betulaceae. In Fl. Reipubl. Popularis Sin. 21: 44--137. Science Press, Beijing.

Schneider, C. K. 1912. Illustriertes Handbuch der Laubholzkunde. Vol. 2. Gustav Fischer, Jena, Germany.

------. 1916. Betulaceae. In C. S. Sargent, ed., Pl. Wilson. 2: 423--508. University Press, Cambridge, Mass.

Smith, W. W. 1915. Diagnoses Specierum. Novarum. Notes Roy. Bot. Gard. Edinburgh 8: 32--33.

Winkler, H. 1904. Betulaceae. In A. Engler, Das Pflanzenreich. 4, 61 (Heft 19). Engelmann, Leipzig.