Taxus Wallichiana Group—Subgroup Wallichiana—Asia

Number of stomata rows/band in leaves of Taxus based on representative herbarium specimens from Asia of the Baccata and Wallichiana Groups.  Note the higher counts for T. contorta (8–11) and lower counts for T. wallichiana (11–14) where the ranges of these species overlap in the central Himalayas.  The Sumatrana Group, which is common in SE China, is not shown; emphasis here is on the relationships between the Wallichiana and Baccata Groups.  The T. cuspidata Alliance includes four species, T. cuspidata in orange, T. biternata in orange, T. caespitosa in red, and T. umbraculifera in black.  The Wallichiana Subgroup includes T. suffnessii in white and T. florinii in blue

     Data for numbers of stomata rows indicate geographical relationships in Asian Taxus are more complex than in North America and Europe; however, they become clearer when other taxonomic features are taken into consideration—as indicated by the species recognized. For example, in the western Himalayas, I recognize T. contorta—where found with 5–8 (-10) or 8–11 stomata rows (Fig. 2C)—by the long narrow leaves that have sclerified parenchyma cells not seen in other species of yew, and by other leaf traits, particularly the elliptical shaped epidermal cells in T-section; collectively, these features indicate the W Himalayan yew is more closely related to yews in Europe than to yews in Asia.  Similarly, in the E Himalayas I recognize T. wallichiana—where found with 11–19 stomata rows—by leaves having angular shaped epidermal cells in x-section, by the persistent bud-scales at the base of branchlets, and by branchlets that show a marked color change in their 2nd yr of growth—from yellowish green to maroon or reddish orange; collectively, the morphological features of T. wallichiana have more in common with yews of SW China than with the W Himalayan T. contorta.

     The Wallichiana (Species) Group is characterized by leaves having reddish epidermal cells and scarcely differentiated stomata bands as seen in dried specimens.  The reddish color, which is most conspicuous on the non-stomatal epidermal areas (adaxial surface, abaxial surface between stomata bands and margins, abaxial midrib), apparently is a product of oil (terpenoid [taxoids] or resin) droplets in cells that become red and deposited on cell walls as leaves dry and deteriorate with age in the herbarium.  Resin in conifers is usually secreted into ducts, but Taxus does not have resin ducts.

     Dried leaves in other species groups usually lack reddish epidermal cells, although similar cells do occur in the mesophyll of the Baccata Group (T. baccata Alliance).  Their non-stomatal epidermal cells—that appear short elliptical in transverse sections—are often colorless, while the color that is evident is due to chloroplast in the underlying parenchyma cells.  In the Baccata Group, leaves appear dark olive to yellowish green with little differentiation in color of stomata bands.  The Sumatrana Group, which shows more differentiation of color, has blood red to yellowish green marginal and midrib areas in contrast to stomata bands appearing olive green to yellowish orange in chloroplast, yellowish orange on cell walls, and yellowish green on papillae.  The color differences between these three groups are obviously chemically related, but despite the wealth of chemical data published on the genus, studies lack authoritative taxonomic documentation.

     Two subgroups of species are recognized in the Wallichiana Group—by the shape of leaf epidermal cells in transverse section.  The Wallichiana Subgroup—characterized by taller than wide to nearly isodiametric epidermal cells in transverse section—ranges from the E Himalayas to  Yunnan, Sichuan, and W Hubei at elevations between 2000 and 3700 m, and is also found in North America from near sea level to above 2000 m.  The Chinensis Subgroup, which has ±elliptical epidermal cells in transverse section, occurs mostly in central China between 300 and 2500 m in elevation, but extends to Nepal, The Philippines and to Indonesia where found mostly above 2000 m.  The Chinensis Subgroup includes plants with rectangular to nearly quadrangular epidermal cells if most other characteristic features of the subgroup are evident such as the vestigial bud-scales at base of branchlets, yellowish tinted (ochre) branchlets, oblong leaves, leaf stomata aligned in transverse rows, and a leaf midrib that is leaves partially papillose.

     The Wallichiana Subgroup is regarded ancestral to other taxonomic groups.  An ancestral species might be characterized as appearing columnar in habit, having conspicuously nerved bud-scales that persist at base of branchlets (T. suffnessii), radial spreading leaves as in T. fastigiata (Baccata Group), long), having tall angular leaf epidermal cells and having leaf stomata in rows that extend across the entire abaxial surface as in T. suffnessii.

     Except for SW China endemic, T. florinii, the North American taxa are more narrowly defined with fewer stomata, less than 12 stomata rows per band (Spjut, 1992, 1993, 1998a, 1998b, 2000c) in which stomata vary from (7-) 8–9 (-11) stomata rows in Taxus globosa var. globosa, from (5-) 6–7 (-8) rows in T. globosa var. floridana, and from 4–7 (-9) rows in T. brevifolia.    In contrast, Taxus wallichiana,  occurring from the E Himalayas to SW China, may have anywhere from 11–19 (-21) rows, or stomata in  T. florinii in SW China may range from 7–12 rows per band.  The leaf anatomical features are given more taxonomic weight in the Wallichiana Group than in other species groups.

IA. Wallichiana Subgroup Asia

1. Taxus wallichiana

2. Taxus florinii Spjut, J. Bot. Res. Inst. Texas 1(1): 222. 2007. —Type: China. Yunnan: Litiping between Likiang and Weihai, tree 17 ft, in mixed forests by stream, R.C. Ching 21980, holotype: A! (Fig. 193, stomata 10 rows/band, abaxial marginal zone 5–6 cells across, these thick-walled).

Florin yew. Distribution: Forest or pasture or along streams, 2500–3700 m; China (Xinjiang Uygur, Sichuan, Yunnan).

Shrub or tree to 13 m high; young branchlets yellowish green, abruptly changing to reddish purple in 2nd yr; bud-scales persistent on 1–2 yr branchlets in most specimens, 2–3 seriate, deltoid, concave, ca. 0.3–0.5 mm long, obscurely to prominently carinate, closely attached. Leaves arranged ±distichous, homomallous, linear-lanceolate, acuminate, straight to falcate, 1.5–3.5 (-4.7) cm long, 2.0–4.0 mm wide, 200–350 µm in thick, dark glossy green above (dried leaves, similarly noted by collectors on fresh specimens), glaucous to yellowish green below (in dried specimens, also noted as yellow green in field), slightly convex above (adaxial) to a rounded midrib, channeled along each side below mid region to base, less concave below (abaxial) to a flush to slightly rounded midrib, margins plane to slightly revolute; upper (adaxial) epidermal cells nearly quadrate in transverse section, occasionally taller than wide, 20–50 µm tall, 20–40 µm wide; lower (abaxial) epidermal cells narrower and not as tall as upper along midrib and marginal regions, rounded to partly angular in transverse section, 5–12 µm tall, 15–25 µm wide, numbering 16–20 across marginal region, trapezoidal to short rectangular, 1–3× l/w near margins, longer nearer stomata bands, 3–5× l/w, and on midrib, 5–10×+ l/w; papillae lacking near margins on 2–10 cells across, positioned more marginally than medially on midrib in 1–2 alternating rows, medially to marginally on accessory cells, and along cell walls in marginal zone; stomata bands narrower than marginal zones; stomata in 7–12 rows/band. Male cones maturing on 1st and 2nd yr twigs, subcylindrical in bud, 3 mm wide, 4 mm long, yellowish green, the scales 4–5 seriate; microsporangia 5–6 on each sporophyll, pinkish, spongy. Female cones initiating on current season growth, subcylindric in bud, 2–4 mm long, maturing 1st and 2nd yr, or on much older branchlets in one specimen, the scales 5–6 seriate; seed ovoid to globose, to 5 mm long, 5 mm diam., pale (yellowish) in color, sharply pointed at apex, covered in part by red aril, Aug.

Representative Specimens—China—Sichuan: handwriting illegible, 3500 m, Schneider 1429 (K); Ngaitaschekou trans flumen Yalung ad septentr oppodi. Yenyüen, 28º15’N, 2800–3500 m, Handel-Mazzetti 2602 (K); Shalie Valley on Muzhiyan Shan above Old Muli, Muli country, ca. 2980 m, frequent in narrow, steep SW facing gulley, tree to 5 m, leaves dark green above, yellowish green below, Fliegner et al. 1129 (K). Xinjiang Uygur (“Soc. W. Sikiang”): “NW Yunnan,” Tamichung, tree 20–30 ft, in fr., open pasture, R.C. Ching 21505 (A). Yunnan: NW: Mt. Fo Ludu, NW of Li-Kiang, W of the Yangtze, J.F. Rock 18502 (A, US); mostly illegible, 3000–3200 m, C. Schneider 2918 (A, K); no locality data, T.T. Yu 11076 (A); Dokerla, 3100 m, Yu 7848 (BM); Litiping Range, Mekong-Yangtze divide, E of Weihsi, J.F. Rock 11573 (A, US); Sikang, Me-kong, Tsa-wa-rung, 2500 m, C. W. Wang 65475 (A); Mekong-Salween Divide, Forrest 19967 (S: C-2084); S Chungtien, Kung-shiang-shu, Snow Mt on the way to Kai- Lou-wei, on Yangtze bank, 2700 m, by stream in wooded side of valley, tree 20 ft, K.M. Feng 3235 (A); same locality, 3200 m, in mixed forest, shrub 15 ft, in fruit, K.M. Feng 1809 (A); Wei-si Hsien, 2500 m, forest, C. W. Wang 67735 (A); Zhongdian, Haba Shan, 27º22'28 N, 100º05'50 E., 3347 m, Alpine Garden Soc. Exped. 309 (K).

     The Florin yew is a polymorphic species with a geographic distribution restricted to mostly Sichuan and Yunnan.  It differs from T. wallichiana primarily by the relatively narrow stomata bands on leaves, usually with 7–12 stomata rows per band; however, stomata density in this species appears correlated with other developmental features such as the width of the stomata band relative to that of the abaxial marginal region, and differentiation of marginal cells by width and thickness of cell walls.   These features vary most in specimens from SW China.  Further study is needed to determine if some specimens listed under T. florinii could be assigned to T. wallichiana var. wallichiana by giving more taxonomic weight to the number of marginal cells lacking papillae, while the number of stomata rows is currently regarded as a practical solution for separating T. florinii from related species.  Additionally, T. florinii consistently differs from T. wallichiana var. yunnanensis in having papillae oriented more marginally than medially on abaxial epidermal cells, but this is not easily determined without experience.  Moreover, it appears that these taxa might be more clearly defined by recognizing another variety or species based on leaves appearing more distantly spaced and having strongly discolorous surfaces as seen in one specimen under T. florinii and another under T. wallichaiina var. yunnanensis. 

     Although  T. florinii is distinguished from T. wallichiana by having fewer than 13 stomata rows per band, the distinction from North America species is not as clear.   Schneider 2918 from Sichuan, for example, is similar to T. brevifolia in having 7 (–9) leaf stomata rows per band bordered by short broad irregularly shaped marginal cells, while Handel Manzzetti 2602—also from Sichuan—with 11 leaf stomata rows per band—is not easily distinguished from T. globosa.  Thus, T. brevifolia and T. florinii could be considered synonymous with the America T. globosa.

     The American species generally differ in the pale yellowish orange branchlets and leaves appearing convex across the adaxial surface, in contrast to the dark purplish branchlets and nearly plane leaves of T. florinii.  Other differences require transverse leaf sections in which the non-stomatal epidermal cells of T. florinii appear smaller with thicker-walls and less sharply angled on the abaxial surface than on the adaxial surface; in the American species, the epidermal cells appear similar in size on both surfaces.  Again, more study is needed to better define the distinguishing characteristics between these species.

     Alternatively, one may include the aforementioned yews of Americas and Himalayas (to SW China) under T. wallichiana.  Such a broad species concept, however, would be difficult to accept from a biogeographical view.  Essentially, characteristics of the three species groups in this study could be referred to as species, and those of the subgroups could be interpreted as subspecies, or one may even prefer the single species concept for the genus, one that I find difficult to accept in view of the long evolutionary history in the genus.

     The epithet is in honor of Rudolf Florin who published extensive studies on taxads employing leaf anatomical characteristics.  Despite some controversies surrounding his ideas on taxad evolution—such as proposing that they be treated in a new class, I have found his reviews and data to be of great value.