North American Species of Taxus
Wallichiana Group (Subgroup Wallichiana)

3. Taxus brevifolia Nuttall, N. Amer. Sylv. 3: 86 (1849). Tab. 108 captioned “Taxus occidentalis”, without reference to specimens. T. baccata var. brevifolia (Nutt.) Koehne, Deutsche Dendrol. 6 (1893). T. baccata ssp. brevifolia (Nutt.) Pilger, Planzenreich 4(5): 113 (1903). Type—U.S.A. Oregon: “Columbia woods,” Nuttall s.n.; lectotype BM! (Nuttall Herbarium, annotated “T. occidentalis” —top specimen with mature male cones, shown here in part; designated by Spjut 2007b); isolectotypes K! PH!. This was originally discovered by David Douglas in 1825 (Sargent 1896). Douglas began his journey up the Columbia River at the mouth in April 1824.  However, the lectotype was probably collected by Nuttall along the Columbia River during the spring of 1835.

Taxus boursieri Carrière, Rev. Hort. 1854: 228 (1854), no specimens cited. Type: U.S.A., California, “forests near stream with Thuja, Abies grandis, Pinus lambertiana,” described as an arborescent shrub with slender branches, leaves 1.5–1.8 cm long, glaucous below, original material unknown. Neotype (designated by Spjut 2007b)—California, Siskiyou Co., Shasta Springs, Jepson s.n., Aug. 1894 (US!).

Taxus lindleyana M. A. Murray, Edinburgh New Philos. J. 1:294. 1855. Also, Rev. Hort. sér 4, 4:379. 1855; Lawson ex Carrière, Traité gén. conif. 523 (1855). Type: U.S.A., California, N and E of San Francisco between 40° and 41° latitude [Klamath Region], “along sides of a glen.under the shade of larger trees”; lectotype (designated by Spjut 2007b) from original material of two specimens at Edinburgh—specimen with accession number E00030316 (E!), shown below, annotation Taxus lindleyana, presumably by Murray, additional notation indicating journal publication (possibly R. Jameson), and another notation indicating “probably only a form of T. baccata, perhaps the T. cuspidatus [spelling?] of Sieb. & Zucc.”  Another specimen annotated T. lindleyana was indicated to have been collected from Vancouver, which may have been before his trip to California.

Pacific yew. Distribution: S coastal Alaska to central California—Coast Ranges at least to San Mateo Co. [based on one or two specimens, one collected by Bolander (US) that is questionable and another collected by Stokes (US)], and quite likely further S to Santa Cruz Co., Laguna Creek, and along the W Sierra Nevada at least to Calaveras Co., Big Trees State Park [possibly in Yosemite Natl. Park (Lemmon, yr 1874), and possibly further S to Tulare Co., S Canadian Rockies—British Columbia and W Alberta, S to W Montana. Often in shade of old-growth forests on N slopes, in bottomland forests, occurring also in seral communities along streams and forest margins; generally in mixed evergreen and montane coniferous forests from sea level to 2650 m; most frequent in forests dominated by western hemlock along W Cascades (Franklin & Dyrness 1969), Douglas fir (Eliot 1938] and lowland fir (Bolsinger & Jaramillo 1990), or sitka spruce and western red cedar in the northern range (Taylor 1932; Taylor & Taylor 1980).

Comments on the southern limits of the geographical distribution (Oct 2007, Jan 2009):  Spjut (1977, 2007) indicated that the southern distributional limits of the species as possibly in Marin County along the Coast and Tulare County in the Sierra Nevada; however, these reports have been questioned. 

The recently published Marin Flora by Howell, Almeda, Follette and Best (Calif. Acad. Sci. and CNPS, 2007) mentioned a discovery of T. brevifolia by Robert H. Menzies in 1932 as being found between Olema and Bear Valley that has since been searched for but not found.   The authors further indicate that Marin County is the southern distribution limits for T. brevifolia along the Pacific coast; however, the southern coastal limits of the species extends to at least San Mateo County based on two herbarium specimens cited by  Spjut (2007a), a US specimen collected by Stokes from San Mateo County, which belongs to var. brevifolia, and another US specimen collected by Bolander (Bolander 186) from “Forest Hills, Devil's Canyon.” 

The locality from which Bolander 186 was obtained is not entirely clear.  “Forest Hills,” “Devil's Canyon,” (Bolander 186, US) has been reported for a location in the Sierra Nevada (M. A. Howe in The Hepaticae and Anthocerotes of California (1899, Mem. Bot. Torrey Club Calif. VII) based on Bolander 4620 for  “Kantia trichomanis”  [=Calypogeia fissa (L) Raddi];  however, all other collections cited by Howe (1899) for this taxon are from the California coast, obtained by other collectors.  Additionally, there is a higher numbered specimen by Bolander of T. brevifolia (Bolander 4554, UC, US) from “near Forest Hills” in Placer County (Consortium of California Herbaria) but without mention of Devil's Canyon (except by Parlatore 1868 without reference to collection number); associated with this number are a series of collections from “Forest Hill” or “Forest Hills (Consortium of California Herbaria),” e.g., Bolander 4551 (Nemophila maculata, “Auburn to Forest Hill at Grizzly Bear House”), Bolander 4552 (Tropidocarpon gracile, “on stage road from Aubern to Forest Hills”), Bolander 4553 (Isopyrum occidentale, “near Forest Hill”), Bolander 4557 (Ceanothus prostratus, “Forest Hill”), Bolander 4558 (Ceanothus tomentosus, “Forest Hills”), and more isolated numbers, 4575, 4576, 4617 and 4629. The Google Earth shows a Devil's Canyon near Foresthill in Placer County.  [A Devil's Canyon also lies further south in the Sierra Nevada off the South Fork Kaweah River just outside the Sequoia National Park in Tulare County, but the vegetation at this site as seen from Google Earth does not appear favorable for T. brevifolia].  There is also a Bolander 186 listed in the Consortium of California Herbaria for Arnica mollis, with locality unknown.  This species is found in the Cascades and Sierra Nevada, including counties of Placer, Mariposa, and Tulare. 

This is in contrast to other collections by Bolander reported from the coast, close to his number 186.  One is Bolander  69 for Carex comosa from the Santa Cruz Mountains (Mackenzie, Monograph of the California species of the genus Carex, 1922, Erythea Vol. 8, edited by Jepson), which a Bolander specimen  without number and location is listed in the Consortium of California Herbaria among only 16 records  for this species, three of which are from the coast, Santa Cruz Co., San Francisco and Sonoma Co. near Guerneville.  Another is Bolander 150 from San Bruno Mountain (which is in San Mateo County) for the lichen Buellia pustullata  (Herre, The lichen flora of the Santa Cruz Mountains, Proc. Wash. Acad. Sci. 12(2): 27–269, 1910).  A gazetteer by David L. Durham (California's Geographic Names: A Gazetteer of Historic and Modern Names of the State, 1998, Quill Drivers Books) indicates a Devil's Canyon extends from San Mateo County to Santa Cruz County [“Peters Creek, stream flows 7 miles to Pescadero Creek 4 miles south of Mindego Hill, lat. 37º15'06"N, long. 122º15 W; sec. 8, T8S R3W, the stream drains Devil's Canyon.  Brown (p. 27), used the name Devil's Canyon Creek for the stream in Devil's Canyon, and noted (p. 68) that the name 'Peters' is for Jean Peter who settled near the creek in 1860”] and that Devil's Canyon itself is in reference to the upper reaches of Peter's Creek in San Mateo County.  A Forest Hill community—that was developed in 1912—is located in Burlingame, San Mateo Co.   It might be noted that in Bolander's Catalogue of the plants growing in the vicinity of San Francisco (1870, A. Roman & Co. Publishers), only Torreya is listed, not Taxus, and that he had also collected in Sonoma County where there is a Forest Hills settlement 4 miles east of Guerneville along the Russian River, while the Consortium of California Herbaria shows only one collection of Torreya by Bolander that he reportedly obtained from Big Trees in Mariposa County.  Here it might be noted that Bolander rarely collected specimens of trees and shrubs; for example, there are no species of manzanita, Douglas fir, or spruce among more than 1200 records of vascular plants in the Consortium of California Herbaria; occasionally he collected a willow or oak or pine.  Notable exceptions include Myrica hartwegii from Big Tree Grove in Mariposa County (Bolander 5002) and Staphylea bolanderi (Shasta Co. McCloud River, Bolander, Kellogg et al. s.n., Apr 1874, type GH) .  Most of his collections appear to be herbaceous species, lichens and bryophytes.

The reported occurrences in the Santa Cruz Mountains are based on John G. Lemmon's Handbook of West-American Cone bearers (1895) which may be traced to C. L. Anderson  (1891; Catalogue of flowering plants and ferns of Santa Cruz County, California; in E. S. Harrison, History of Santa Cruz County, California, Pacific Press Publ., San Francisco, Chapter IX).  Anderson (1891) made specific reference to Laguna Creek for T. brevifolia, suggesting a rather localized occurrence in Santa Cruz County, compared to other comments he made about the occurrences of plants in his checklist.  This location with reference to C. L. Anderson was also mentioned by Jepson and Dempster in A flora of California (1909, Cunningham, Curtiss & Welsh). Modern checklists of the area (e.g., R. Morgan et al. 2005.  An annotated checklist of the vascular plants of Santa Cruz County, California.  California Native Plant Society; SEINet 2006 (previously available online),  did not include T. brevifolia, however.  Dr. Anderson, a physician, frequently sent his specimens to Asa Gray for identifications, but there are no specimens of T. brevifolia from the Santa Cruz Mountains in Harvard and other California herbaria (Thomas, Flora of the Santa Cruz Mountains of California, 1991, Stanford Univ. Press), while it must be noted that Anderson's (1891) checklist was derived largely from another by F. L. Clarke (1890, Catalogue of flowering plants and ferns of Santa Cruz County, California, Pamph., 13 pp.; see Bull. Torrey Bot. Club, 1892) based on specimens collected by students as a result of a county-wide “wildflower contest” among the public schools.

[Updated 22 Dec 2009]. The southern distribution limits of Taxus brevifolia in the Sierra Nevada has been questioned (P. W. Rundel, The southern limits of Taxus brevifolia in the Sierra Nevada, Madroño 19: 300, 1968; Spjut 1977). It has been reported as Tulare County (Jepson, Silva of California, 1910; Univ. Press, pp. 164-168; Sargent, Silva of North America 1896) with further reference to Sequoia National Park (Spjut 2007b); however, present day records of the species are not known south of Calaveras County (the North Grove, Calaveras Big Trees State Park. Dean Wm. Taylor, pers. comm., Sep 2007 (see also http://californiabotany.blogspot.com/2009_04_01_archive.html); Rundel 1968).  Other reported occurrences south of Calaveras County are based on specimens from Mariposa County collected by J. W. Congdon (15 Aug 1901, E00094392, E00094387) and by Lemmon (John Muir) from Yosemite Natl. Park (UC334344; US; collected Sep 1874) and literature with references to Merced Canyon (K. Brandegee, The Flora of Yo Semite, Zoe  2: 155–167 , 1891; also in Harvey and Carlotta Hall, A Yosemite  Flora, 1912).  It is interesting to note that Lemmon (1895) reported only Torreya in Yosemite, while there are no Lemmon specimens of this species from Yosemite in the Consortium of California Herbaria, only Taxus.  Additionally, a Natureserve report (Classification of the Vegetation of Yosemite National Park and Surrounding Environs in Tuolumne, Mariposa, Madera and Mono Counties, California, 2002) mentions T. brevifolia as infrequent, occurring with Torreya in  a Pseudotsuga menziesii - Abies concolor - Calocedrus decurrens association (updated June 2010). Additionally, T. brevifolia may have once occurred in the southern Sierras—on the Kern Plateau—during the Quaternary based on pollen studies (Axelrod in Barbour & Major, Terrestrial Vegetation of California, 1977, John Wiley & Sons, NY with reference to D. I. Axelrod & W. S. Ting, Early Pleistocene  floras from the Chagoopa surface, southern Sierra Nevada. Univ. Calif. Publ Geol. Sci 39: 119–194, 1961.).

General Comments on the Species, T. brevifolia. This species is distinguished by its leaves having relatively large angular epidermal cells as seen in T-section (see below),  by relatively narrow stomata bands with usually 4–7 stomata rows, and by the trapezoidal or wedge-shaped cells in the abaxial marginal zone (see illus. in key). The leaf arrangement, shape and texture, and large angular cells in T-section, and the persistent bud-scales are features also seen in a rare Himalayan species, T. suffnessii Spjut (Myanmar), one that I regard as having most of the ancestral features of the Wallichiana group.  Nevertheless, leaf anatomical data suggests T. brevifolia was derived from T. globosa by loss of stomata and increase in cell size, possibly as it adapted to a drier climate that developed during the Tertiary.  Moreover, I believe that Mesoamerican yew  descended directly from the Suffness yew, and that the Florin yew, although nearly indistinguishable,  evolved independently.

Although T. brevifolia is typically a tree, four rather distinct shrub forms can be recognized in the field; however, I have not been able to distinguish them in the herbarium with exception to possibly one that occurs in the Klamath  and Rocky Mountains (Spjut et al. 1993).  Three varieties have been described based on whether plants reproduce by layering, and by differences in length and persistence of cone scales have been recognized (Spjut 2007b), and a fourth variety is also proposed (Nov 2011) based on habit of the plant appearing similar to the Irish yew .

3a. Variety brevifolia. Diagnostic features: Tree with erect trunk at base, reproducing vegetatively by adventitious shoots from trunks or roots; branches developing in all directions around the trunk, wide spreading to ascending or somewhat erect;  ovuliferous shoots usually with one ovule, rarely two or three; seed longer than the primary shoot.

Tree, reproducing by adventitious shoots from trunks or roots, 6–13 (-25) m tall, bole to 60 (-130) cm diam.; branches horizontal to upwardly ascending, sometimes drooping, dividing more unequally than equally, yellowish orange to reddish orange, or maroon; bud scales persistent on 2^nd–3 ^rd yr branchlets, conspicuous, 2–3 seriate, paleaceous, chartaceous, closely to loosely adnate, plane to slightly concave, indurate, brownish, 1–1.5 (-2) mm long. Leaves spreading nearly in one plane and diverging along opposite sides (±2 ranked), overlapping, ±heteromallous, oblong to linear, straight to recurved and slightly falcate, (1.0-) 1.4–2.9 cm long, 1.0–3.5 mm wide, 200–350 µm thick, obtuse to acute at apex, glossy green and convex above to a rounded or acutely keeled midrib, yellowish, dark green or glaucous and concave below to a rounded midrib, revolute when dried, especially in upper third. Adaxial epidermal cells often taller than wide, rectangular, 25–55 µm tall, 25–40 µm wide, strongly mamillose; abaxial marginal epidermal cells similar, irregular in width, trapezoidal, thin-walled and inflated in part, mostly 1–5× l/w, or 5–10× l/w in northern plants, papillose on nearly ¼–½ or more of the marginal cells—extending to 2 rows of cells from margins, papillose across midrib except near base, or rarely less papillose in center rows; papillae in 2–4 rows on each cell, positioned more marginally than medially, along cell walls; stomata mostly in 4–6 irregular rows/band, rarely in 7–9 rows/band.  Cones (fertile short shoots) one per leaf axil, or occasionally two arising in the axils of leaves with one more dominant (larger); scales forming a basal hemispherical cup; male cones ellipsoidal in bud, 4 mm long, 2 mm wide; sporangia exserted.  Ovuliferous shoots with one terminal ovule, occasionally two or three with one that is more dominant than others, or additional ovule found near the base of the ovular shoot complex that may be partly in the axis among the basal scales without any clear association with a scale. Seed on 1^st or 2^nd yr or older branchlets, longer than the primary shoot, rounded, slightly 2–4 angular, ellipsoid to ovoid, tapering to apex from mid region, or rather abruptly near apex, 5–8 mm long, to 4 mm diam.; aril red, reddish orange, yellowish orange, or rarely yellow, Aug.-Sep.

Representative Specimens—U.S.A.—Idaho: Idaho Co.: near Allison Creek, bridge 221-65.1, T28N R2E Sec 14, tree 28' high, 15.8" dbh (fresh), Shields s.n. (wba). Washington:  Lewis Co., Gifford Pinchot Natl. For., 7 mi N of Packwood along State Hwy. 5, small tree to 8 m, Barclay et al. 1645, voucher for antitumor active sample in discovery of taxol (NA); same locality, tree to 35 ft, Barclay & Arguelles 1771, voucher for recollection from which taxol was isolated (Wani et al. 1971) (NA). Oregon: Clackamas Co., Mt. Hood Natl. For., above Ripplebrook For. Sta., Perdue 13002 (NA); W Cascades, Clackamas River area, T7S R8E, tree number 007-011, 930 m (fresh), Lankford s.n. (wba); Josephine Co., ⅛ mi S on road 2500.052, above Taylor Creek on Minnow Creek Road, T35S R8W Sec. 27, 650 m, USFS s.n. (wba); same locality, just above jct. with road 25, tree form, 70 cm to 10 m tall (fresh); T35S R8W Sec. 34, tree, 16.3" dbh (wba); Columbia River, 1860, Lyall s.n. (K). California: Del Norte Co., Oregon Mt. Rd., 1 mi up from S end, along both sides of road, T 18N R4E Sec. 16, 200 m, fresh specimens from various trees and heights, female tree—top middle and bottom branches, lower branch of female tree 12-13 m high and 50 cm diam., middle branch of female tree 6-7 m high and 15 cm diam., middle branch of female tree 5 m high and 7-8 cm diam.; top branch of male tree, lower branch of male tree 13 m high and 30 cm dbh, Yager & Bozovsky s.n. (wba); Siskiyou Co., Marble Mts. Wilderness Area, Lovers Camp, 2000 m, tree to 10 m (fresh), Spjut 10171 (wba); Salmon/Trinity Mts., Klamath Natl. For., 11 mi W of Callahan-Etna Road along French Creek Road, Duck Creek drainage, mixed evergreen forest in narrow ravine. 41º20' N, 122º54.32'W, 3700 ft, tree, 4–6 m high, Spjut 12309 (wba); Trinity Co., Slate Creek near Tannery Gulch, Trinity Lake, 700 m, Spjut 10179 (wba); Humboldt Co., base of heavily wooded slope bordering flood plain of E fork of Willow Creek, ca 6 mi W of Willow Creek town, near Hwy. 299, 1500 ft; same locality, Terrell et al. 4170 (NA).

 

Taxus brevifolia.  Oregon—Type (BM).  Lectotype: upper left specimen.	Taxus brevifolia var. brevifolia.  Washington.  Thurston Co., Barclay 1645, Aug 1962; voucher for sample that led to the discovery of taxol, photo showing close-up of seed.
California. Humboldt Co.  Inner Coast Ranges.  Terrell et al. 4170, sample from one tree, special study.	Oregon: Near Ashland. Area clear-cut, except for yew trees.
Collection site near Terrell et al. 1970 shown above. Photos by R. Spjut taken July 2010 (Spjut 16716).  Left: habit of Taxus brevifolia var. brevifolia.  Right: close-up of branch with two ovular shoots in the axil of one leaf, one evidently aborting development; another shoot with one ovule where two appears to have been present.  The upper pair of ovules may have initially developed terminally among the bud-scales—see var. reptaneta below with terminal male strobilus and terminal female cones.  A vegetative or fertile shoot appears to have subsequently continued to grow.	Marble Mountains, Lovers Camp Trailhead to Wilderness area, Spjut 10171, July 1987.  Area evidently selectively logged in the past, yew trees left standing.
California: Shasta-Trinity NF, Trinity Co., Stuart Fork Valley, July 1997. Area clear-logged and yew trees left, the foliage turned orange from exposure to sun.  Tree in forest on steep bank.  Close-up of branchlet with male cones.	California.  Klamath NF.  Callahan-Cecilville Rd, 2900 ft, Douglas fir, ponderosa pine forest, Spjut 16019, Aug. 2006. Photos show trees with both single and multiple trunks from base.
California. Klamath NF: Siskiyou Mts. NW of Yreka; Rich Spjut & Susan Spjut 15802, Oct. 2004.	California—Var. brevifolia, showing mostly plants developing from what appears to be a burl-like base.  Various plants photographed along the North Fork of the Salmon River, Marble Mt Wilderness, between 3000 and 3500 ft., July 2006.
Taxus brevifolia var. brevifolia.  Specimens with seed.  Left: from California, Sierra Nevada, Yosemite, Lemmon, yr 1874 (US).  Right: California: Mt. Shasta, Brewer 1361 (US).	Taxus brevifolia var. brevifolia.  Specimens with male cones.  Left: California. Sacramento Valley, USDA without number, yr 1888 (US). Right: California. Sierra Nevada, Placer Co. Bolander 4554 (US).
Taxus brevifolia var. brevifolia.  California.  San Mateo Co. Stokes s.n. (US).	Taxus brevifolia var. brevifolia.  California.  Mendocino Co., Clark s.n., yr 1877 (US).

3b. Taxus brevifolia var. polychaeta Spjut—J. Bot. Res. Inst. Texas 1 (1): 217. 2007. Type: U.S.A., Washington: Thurston Co., Mud Bay, near Tacoma, F. G. Meyer 1589 (holotype: K!), photo shown in key.  Note: A combination attributed to Silba, Taxus brevifolia subsp. polychaeta  with reference to "J. Int. Conifer Preserv. Soc. 17(1): 22. 2010" (n.v.) is inconsistent with taxonomic treatment of subtaxa within the genus (Silba 1984; Spjut 2007a, Spjut 2007b).  Subspecies status in Taxus was applied to geographically differentiated populations by Pilger (1903, 1916, 1926).  Thus, Silba's proposed combination would appear to be superfluous, and possibly invalid. 

Diagnostic features: Tree with erect trunk at base, vegetative reproduction unknown; branches developing in all directions around the trunk, wide spreading to ascending or somewhat erect; ovuliferous shoots with numerous scales that are all similar in size (belonging to the primary shoot), much longer than seed, sometimes shortly branched near base, or branched above base (Sonoma County, CA), usually with more than one ovule, two subterminal ovules on long primary shoots , 5–8 mm long, in the northern range, or as many five ovules on branched shoot complexes <5 mm long in plants from Sonoma County, CA.

Worm-cone yew: California (Mendocino Co.?, Sonoma Co.), Washington (Mud Bay near Tacoma, extirpated?).  Specimens studied: California: Sonoma Co.: 7 km E of Stewarts Point, redwood-grand fir forest, 200 m, Rich Spjut & Rick Spjut 16021 (wba). Bigelow (Marin, Sonoma or Mendocino Co.), yr 1854 (US); Salmon Creek (Sonoma or Mendocino Co.?), McMurphy 315 (US).  Idaho, near Coeur d'Alene, shipped by Marion Blatch as fresh material via overnight mail from Coeur d'Alene Nursery, 26 June 1992, possibly collected in nearby Washington; collector, locality, and habit of plant unknown, supplied by the USDA Forest Service without data.  A report by Spjut (1977) noted a collection from “Coeur d'Alene National Forest: T51N R8W Sec. 25; 3,800 ft, on north slope near stream” (n.v.).

This new variety has been recognized by elongate cones (well-developed primary fertile shoots) that resemble annelid or polychaete worms, hence, the epithet, polychaeta (Spjut 2007).  Variety polychaeta apparently occurs with the typical variety (var. brevifolia), judging from other specimens—collected by Fred Meyer reportedly at the type locality under the same number (e.g., specimen at US)—that have typical cones.  While Spjut (2007) had  earlier questioned whether var. polychaeta  has biological 'species' significance (Spjut 2007), he also noted that unique genotypes in T. brevifolia have been recognized in relationship to other geographically differentiated genotypic populations (El- Kassaby & Yanchuk 1994), but without consideration to morphological and reproductive differences (Spjut 2007). 

The ovule bearing shoots (cones) of var. polychaeta were recently reported distinct from those of other species and varieties by their development of multiple ovules (this web page, June 2010); however, vars. brevifolia and reptaneta have since been found to have multi-ovulate shoot complexes.  The ovules may appear terminal on a ovular shoot complex (var. brevifolia, polychaeta, reptaneta), or in var. polychaeta the secondary shoot may further grow from an aborted development of one of the ovules, leaving the lower ovular region to appear lateral with one or more terminal ovules, or as others have suggested, multiple secondary shoots may arise from a primary shoot over a period of several seasons (Anderson & Owens 1999; Anderson 2001).  It has not been determined whether the primary shoot falls with the seed or persists in var. polychaeta.

The idea of Taxus having just one terminal ovule on a shoot that arises from the axis of a leaf has generally prevailed in the literature on Taxus reproductive morphology, and because there is no clear homologue to the seed-scale-bract complex in conifers (e.g., Florin 1948, 1951; Cope 1998), Taxus is also stated to lack cones; however, the definition of cone given by Clement-Westerhof (1988) includes the “ovuliferous dwarf-shoot” as it relates to extinct conifers, and further defines bract as “a foliar appendage of the cone axis.”   It should be recognized that the ovular shoot complex of Taxus consists of a primary shoot that arises in the axil of leaf and a secondary shoot that arises in the axil of a scale.  The interpretation of whether cones are present or absent in Taxus may be a question of whether it is the subtending scale of the secondary shoot, or the leaf that subtends the primary shoot is the bract homologue (Miller 1988).  Examples of various multi-ovulate shoots of Taxus are presented elsewhere—in an overview of the genus Taxus.

The occurrence of var. polychaeta along the California coastal region is rare (Baye 2004, ltr to Calif. Dept. Forestry; Spjut 2007).  One record cited for Sonoma County by Spjut (2007) was from Annapolis (Milo Baker s.n.).  In a brief survey of this area by Rich Spjut and Rick Spjut in August 2006,  we found only one tree (Rich Spjut & Rick Spjut 16021; Spjut 2007).

However, additional plants have been found by Peter Baye at widely scattered spots along Fuller Creek and its north-facing drainages as noted by him in a report (letter) to the California Dept. of Forestry (www.gualalariver.org/vineyards/martin-baye.pdf), and subsequently has found it at other locations (Peter Baye, pers. comm., Mar-June 2010), some of which have been studied by Spjut (Spjut & Baye proposed publ. on Taxus in Sonoma County).  Generally, plants of var. polychaeta occur as isolated individuals at intervals of 300–500 m, or may be found in small numbers of up to three or more plants within 50 m of each other.   Nonetheless, the occurrences are still considered  rare, especially in view of the narrow ecological zone in which they seem to occupy (Spjut & Baye proposed publ.).  Peter Baye has also reported finding plants of var. brevifolia.

Other coastal occurrences were reported from a ravine on Sea Ranch in southern Mendocino County, an area reportedly logged in early 1900’s and again in the early 1990’s (Web report previously cited but no longer available online, or has changed), and also one specimen provided by the participants in the California Consortium of Herbaria collected from near the town of  Mendocino; ucjeps.berkeley.edu/consortium/; Fri Sep 3 14:49:50 2010; Wheeler & Gladys L. Smith 2681 (collected 29 Jul 1981), HSC83650.     

It was reported that the specimen with a male cone (strobilus) collected by John Milton Bigelow from California corresponded to a male plant of var. polychaeta (Spjut 2007, also shown above).  However, male plants in the general vicinity where female plants have been found of var. polychaeta appear similar to male cones of var. brevifolia, except for possibly being smaller in size.   Bigelow is known to have collected in Marin, Sonoma and Mendocino Counties (CNPS, Marin Chapter, website).  More field studies are needed during the spring, along with comparative DNA and morphological analyses.  This may help clarify whether more than one variety or species of Taxus occurs in Sonoma County. 

 3c. Taxus brevifolia var. reptaneta Spjut, J. Bot. Res. Inst. Texas 1 (1): 219. 2007— Type: U.S.A., California, near corner of Humboldt, Siskiyou and Trinity Counties, 17 mi S of the town of the Forks of the Salmon on McNeal Creek Road, Knownothing Creek, ca. 1300 m, shrub with long scandent stems ascending to 5 m or more, stems layering, R. Spjut & T. Spjut 11835, with seed—holotype US (isotypes: BM BRIT E GH HSC K). Note: A combination attributed to Silba, Taxus brevifolia subsp. reptaneta  with reference to "J. Int. Conifer Preserv. Soc. 17(1): 22. 2010" (n.v.) is not consistent with the taxonomic history of the genus as noted above under var. polychaeta.

Diagnostic features: Shrubs with ascending stems near base, or ascending from a prostrate trunk; vegetative reproduction often by layering, forming clonal thickets or genets with many ramets; branches developing in one plane along a stem (often appearing two-ranked), often spreading in one direction (appearing falcate-secund), ascending and usually with a drooping leader; cone scales persistent; ovuliferous shoots relatively short as in var. brevifolia but not uncommonly found with two terminal ovules—i.e., in ~25% of the female genets (based on limited surveys in 2010).

Thicket yew. NW U.S.A. Scattered or forming dense thickets on steep sunny slopes of avalanche shoots, along streams, or in dense shade of bottomland ravines, 1000–2000 m; Rocky Mountains—Idaho, Montana, British Columbia; E Cascades—Oregon and Washington, Klamath Mountains—Oregon and California; Coast Ranges—Oregon.

The layering habit of var. reptaneta varies. The variety at the type locality and in NW Montana, which occurs in open riparian habitats, has thick trunks—more than 50 cm in diam.—that ascend to 5 m or more in length. These bear numerous divaricate to ascending branches sometmes longer than that of the main trunk.  New trunks develop from layering branches within less than 5 m. Collectively, they form impenetrable yew thickets—as I have observed on sun-exposed avalanche shoots in the Kootenai National Forest of NW Montana, and on steep slopes bordering streams in the Klamath Mountains of California (type locality); hence, the epithet, reptaneta. Another form, which occurs in understory of old-growth bottomland (valley) forests along the E slopes of the Cascades, has more scattered genets, and their trunks sometimes twist and wrap around the base of grand fir and other trees as if ecologically dependent upon them. Another form of a much lower stature hardly differs from that of T. canadensis, except for its leaf anatomy; it is common in the Rocky Mountains.

 

California: Klamath National Forest, Marble Mountain Wilderness, Lake of the Island, 5,650–6,000 ft.  Arrows in upper left photo point to yew thickets on north to northeast exposures, occurring on almost vertical rock faces and steep talus slopes, generally of diorite; Spjut 16013, July 2006.  Upper right photo shows closer view of yew thicket.  Lower left photo shows tangled growth of branches, which appear difficult to walk through.  Lower photos show male cones found on all plants examined in one thicket, as would be expected from plants that layer.  Note the trailing stems along ground in lower left photo.  Occurring with red fir, mountain hemlock, white fir, silver fir, western white pine, incense cedar, and subalpine fir (very rare).  Douglas fir occurring just below the lake along the trail at 5,400 ft. An orange color appearing on older leaves of many plants is an apparent disease that I have not generally seen on yew.	California: Marble Mountain Wilderness, Lake of the Island, small thicket along north shore, 5650 ft.  Here the ground is level, but the plants still show the ascending stems as if growing on steep slopes.  Plants in this thicket were entirely female in contrast to male thickets seen elsewhere around the lake. Spjut 16015, July 2006.	California: Marble Mountain Wilderness, Lake of the Island, small thicket along east shore, 5800 ft. Spjut 16014, July 2006.
California: Marble Mountain Wilderness, North Fork of the Salmon River just below junction with Lake of the Island Trail, 4000 ft. Plants forming thickets but appearing intermediate in characters of phyllotaxy and habit to var. brevifolia.  These plants have less crowded leaves in a more definite 2-ranked arrangement, and the basal branches were not as trailing.  This form is similar to plants seen in understory forests in the eastern Cascades of Oregon and Washington; Spjut 16016, July 2006	Idaho: Hobo Cedar Grove Botanical Area, Taxus brevifolia var. reptaneta, a predominant understory shrub in an old growth forest of largely western red cedar. Photo shows genet with ascending ramet stems.  Note the dominance of branches spreading to the right side of the ramet stems.	Type locality, Knownothing Creek, 17 mi S of the Forks-of-the-Salmon town.  Left photo shows yew thicket.  Aug. 1991. Male cone, a strobilus and bud-scale-dwarf-shoot and leaf complex at the end of a primary (fertile) shoot of Taxus brevifolia var. reptaneta from type locality (Spjut 16717). Photo taken July 2010.
Washington—Cascades: Washington National Forest, 18 mi S of Leavenworth, Icicle Creek, Black Pine Horse Camp, forming thickets locally on flats or gentle slopes in understory of old growth coniferous forest with Engelmann spruce,  mature Douglas fir forest, incense cedar, grand fir, sugar pine and ponderosa pine, 47º36 N, 120º56.70W, 3050 ft, Spjut 12302, 12101.  In this area, thicket yew seemed to show a definite close association with other conifers by growing around their base as evident in the above photos. Lower right photo shows yellowish seed attached to branch.  Lichen growth on yew branches indicate that yew plants are not young.  Three photos of Taxus brevifolia var. reptaneta from Icicle Creek taken July 2010 (Spjut 16723).  Left:  habit of genet (in part) with ascending ramet stems, which have long divaricate branches, all pointing to one side of the stem.  Center: two fertile branches, one with a bi-ovulate cone, another with two ovules arising in the axils of successive leaves. Right: A short primary (fertile) shoot terminating in two cones, one apparently aborting (could also become a vegetative shoot), and the other bearing two ovules.	Canada—British Columbia:  Between Rocky Mountain National Park and Glacier National Park, 2.5 mi W of Golden, S off Hwy 1, S of Big Lake Resort, elev. ca 1000 m.  On steep rocky slopes of narrow ravine, occurring with Douglas fir, western red cedar, maple and hazelnut, Spjut 12090.  Thought to be distinct for the basal burl, but this character is not easily evaluated without causing damage to plants.	NW Montana.  Lincoln Co., near northern panhandle of Idaho, Kootenai National Forest, Libby Mt., Snowshoe Mine Road, T28N R31W Sec. 495, Top photos— provided by David Deevy—shows thicket of yew before collectors debarked stems.  Right photo shows a good view of the habit of the plant after collectors removed branches and bark from plants.
Oregon: E Cascades, E of Portland, Spjut 12301, Aug. 1992.	Montana: Flat Head NF, Swan Lake Dist. Big Head, Bay Creek road (FS Road 129), T25N R 18W Sec 6; Spjut & Broster 12308. Aug 1992.  Showing low creeping habit much like T. canadensis, also in the vicinity one large tree of typical T. brevifolia.

I had considered thicket yew as a distinct species by the darker green foliage with leaves overlapping more closely and irregularly compared to the typical tree form.  This was based on plants in the Klamath Region of California.  It was even possible to recognize herbarium specimens that had these features, but primarily in California.  A good example was a specimen I recognized in the HSC herbarium, Oettinger 504 (MA thesis, 1975); it was reportedly collected in the vicinity of English Peak, northeast side of Hancock Lake, in montane chaparral, an open dry site of decomposed granitic rock, 2000 m, further indicated as one shrub, 4 ft high and 8 ft in diam. (HSC).  This is not far from the Lake-of-the-Island site shown above. A well-known packer in Salyer CA, Dean McCroom, reported (pers. comm. July 2006) that thicket yew is quite common on north to northeast exposures of English Peak near  6,000 ft. in elevation. The English Peak area is know for other relict conifers.

In studies with assistance from the USDA Forest Service in the Pacific Northwest Region, the author was directed to many other sites during Aug 1992 where "shrub yew" was known.  At least four distinct growth forms were recognized in the Klamath Region, Eastern Cascades, and Rocky Mountains.  But these forms could only be distinguished in the field, not from herbarium material alone.  Nevertheless, each seemed worthy of taxonomic status.  Samples from each of these and also of the typical tree variety (var. brevifolia) were obtained for taxoid analysis by Professor C.-j Chang.  The NW Montana collection with David Deevy and samples from the type locality near the Forks-of-the-Salmon in California showed remarkably similar taxoid content, and were clearly distinct from samples of var. brevifolia (Spjut et al. 1993).  Dr. Chang felt that this was sufficient evidence for recognizing it as a distinct species; however,  one cannot simply ignore the other shrub samples; they too had to be dealt with before any new species could be described.  Fifteen years later, var. reptaneta  still appears to be a distinct species.  The problem remains, however, even with support from molecular data, that one should be able distinguish a species in herbarium specimens using morphological characters, and there is also the question of whether all shrub forms belong to a distinct species with many varieties or should be simply treated as a single variety as I have done here.  Nonetheless, the various shrub forms all show the diagnostic features of ascending stems and branches spreading more to one side than equally all around the stem.

Additionally, fertile shoots may have biovulate cones, or two terminal cones where only one bears one or more ovules, or ovules develop in successive leaves so as to appear biovulate as often seen in T. baccata var. dovastoniana.  These findings would seem to indicate that ovuliferous shoots (as well as male strobili) in Taxus are not strictly axillary to a leaf.

 3d. Taxus brevifolia var. nov.

 

Taxus brevifolia var. nov.        Columnar trees similar to the Irish yew in growth form, evidently adapted to growing in open sun exposed habitats, colonial or isolated trees scattered in woodland bordering a lake. Branches dense, developing from near the ground. One tree found with fewer branches near base had bark removed for a short distance up half of the trunk.  This may explain the yellowish green foliage for that individual. Leaf arrangement similar to Taxus umbraculifera. Leaf anatomy indistinguishable from var. reptaneta. Male cones appear in successive leaves near ends of branchlets, maturing later than in other varieties, the above photographed in early August (2011) at 4,800 ft elev. Female cones developing in axils of successive leaves, and at end of primary branchlets; immature seeds notably ellipsoid, not ovoid as generally seen in the species.  The above combination of features suggest a new species; however, the tall rectangular leaf epidermal cells as seen in x section, 7 stomata rows per band , and distribution of papillae across the abaxial leaf surface all agree with T. brevifolia.  Of local and frequent occurrence in the Klamath Region of the Pacific Northwest.  Details of the location not revealed to help protect the plants, which may prove to belong to an extremely rare and endangered yew variety of the Pacific Northwest.

7. Taxus globosa Schlectendahl, Linnaea 12: 496 (1838), no specimens cited. Lectotype (designated by Spjut 2007b): Mexico, Hidalgo: Real del Monte, Cerro de las Nabajas, [Sep] 1837 Ehrenberg s.n. K! (photo at right); isolectotypes L, S, US!. 

Mesoamerican yew. Distribution: El Salvador and Honduras to Florida.

7a. Var. globosa.  Low spreading shrub or tree, to 40 m tall, bole to 40 cm diam.; bud-scales loosely persistent, 2–3 seriate, brownish, 1.5 (-2) mm long, chartaceous, dull, lower scales ovate, upper narrower, plane to slightly concave, carinate towards apex, or cuspidate. Leaves spreading as in T. brevifolia, overlapping in two-ranks, linear-falcate, 1.5–2.2 (-3.5) cm long, 1.5–2.2 (-2.5) mm wide, 200–350 µm thick, pale to dark green and slightly convex above to a rounded midrib that is slightly channeled at base, yellowish or pale green and slightly concave below to a slightly rounded midrib, plane to revolute near margins, acute to apex. Upper (abaxial) epidermal cells isodiametric to slightly elliptical in transverse section, or slightly taller than wide, 20–25 (-40) µm tall, 20–30 (-45) µm wide; lower epidermal cells similar to upper in transverse section, fusiform to rectangular near margins, slightly inflated, narrower and more incrassate towards stomata bands, mostly 3–10×+ l/w, longer on midrib, papillose to 3–4 rows from margins; papillae conspicuous in 2–3 alternate rows on each cell, positioned submarginally and marginally (along cell walls); stomata (7-) 8–9 (-11) rows/band. Male cones not studied; aril red, maturing in the fall.

Mesoamerican yew. Distribution: montane cloud forests above 2,000 m, S Mexico (Veracruz, Hidalgo, Mexico, Oaxaca), Guatemala (Baja Verapaz, El Progreso [Sierra de las Minas, Volcán de Santa Luisa], Zacapa, Volcán Gemelos, Monte Virgen] Huehuetenango [Sierra de los Cuchumatanes, Cerro Cananá]) to British Honduras (Merrendón, San Idalfonso, Omoa, Montecillos, Opalaca, Congolón) and El Salvador.

Representative Specimens—Mexico: Trinidad Iron Works, 24 May 1904, Pringle s.n. (US). Tamaulipas: from El Cielo to Ojo de los Indios and the Parque above Gómez Farías, A. J. Sharp 52112 (GH). Veracruz to Orizaba, 1851, Meisner s.n., Herb. Gordon (K). Hidalgo: barranca below Honey Station, Pringle s.n. (BH); Real del Monte, Hartweg 438 (BM). Guatemala: Dept. El Progreso: hills N of Finca Piamonte, 2400–3333 m, Steyermark 43487 (F); W end of Sierra Las Minas, Baja Verapaz, 6500 ft., Sharp 45247 (F). Honduras: Cerro Sta. Barbara, steep rocky slopes, 2750 m, tree 30–50 ft, usually of low spreading, or even semi-procumbent habit, 10–12" diam. at base, Armour & Chable 6083 (US). El Salvador: Dept. Chalatenango, Summit to Los Esesmiles, 14º20'N, 89º09'W, dense cloud forest, 2670 m, Tucker 1073 (BH, PH, US).

This species is recognized by the abaxial leaf epidermis having relatively narrow wavy-walled cells.  Specimens from northern Mexico are not easily distinguished from those in Florida, thus, I conclude that the Florida yew also occurs in Mexico, and that differences between the two varieties are relatively minor compared greater variation in their Asian relatives.  The Florida yew is also difficult to distinguish from related species in SE Asia,  particularly T florinii in Sichuan and Yunnan, and T. phytonii in the Philippines, Taiwan, and Yunnan their epidermal cells differ in being slightly inflated near margins, and by the slightly elliptical shape in T-section. 

7b. Taxus globosa var. floridana (Nutt. ex Chapman), Spjut, J. Bot. Res. Inst. Texas 1(1): 224. Taxus floridana Nuttall, Chapman, Fl. South. U.S. 436 (1860), no specimens cited; T. baccata ssp. floridana (Nutt.) Pilger, Planzenreich 4(5): 113 (1903); T. canadensis Marshall var. floridana (Nutt. ex Chapman) Silba, Phytologia Mem. 7: 72 (1984). Type. U.S.A. Florida: near Aspalaga, 1833, Croom s.n.; lectotype (designated by Spjut 2007b): PH! (with label in Nuttall handwriting).  Specimens referred to as T. croomii, which were obtained by Chapman, are also original material, but are not isolectotypes as earlier noted on this site.

Tree or shrub, 6–10 m tall, bole to 40 cm diam. Leaves mostly (1.0-) 2.0–3.0 cm long, 0.3–1.5 (-2.2) mm wide, 200–350 µm thick, often glossy dark green and convex above to a rounded midrib, yellowish or pale green and concave below to a rounded midrib, recurved near margins. Epidermal cells nearly isodiametric to nearly quadrangular in transverse section, similar in height on both surfaces, 25–40 µm diam., those on abaxial surface uniformly rectangular, thick-walled, larger or slightly inflated or irregular in shape in up to 6 rows nearest stomata bands, mostly 1–5× l/w near margins, more than 10× l/w on midrib; papillae lacking on 5 or more cells across margins, mostly concrescent, globose to cylindrical, in 2–3 linear rows on each cell, positioned marginally, along cell walls, prominent in stomata bands, less developed on midrib; stomata (5-) 6–8 rows. Male cones numerous on young branchlets, scales 3-seriate, microsporophylls ca. 8, each with ca. 6 cucullate microsporangia. Female cones on 1^st yr branchlets, the scales decussate in 3–4 ranks; aril red, Sep.-Oct.; seed conical, 4-angled.

Florida yew. Rare, W Florida: Apalachicola River; Chamaecyparis swamp ca. 8 mi SE of Bristol) (http://www.floridata.com/ref/T/taxu_flo.cfm);  N Mexico (Nuevo Leon, Tamaulipas).

Representative Specimens—Mexico—Veracruz: 2200 m, Hernandez 01459 (BM). Tamaulipas: Sierra Madre Oriental, ca. 30 mi S of Monterrey, C. H. & M.T. Mueller 1337 (BM, PH); Tamaulipas/Nuevo Leon, 24º99'N; 100º5'W, 1690 m, Meyer & Rogers 2746 (BM). U.S.A.—Florida: near Bristol, Mohr s.n. (PH), ex Canby Herb. (PH); Rock Bluff, limestone, Blanton 7050 (PH), Wherry s.n. (PH). Cultivation. University of Florida, Gainesville, hedge like shrub outside campus building, Gilliland & Spjut 12172 (wba); fresh specimen without collector or locality, Snader s.n. (National Cancer Institute). Other material submitted by Phyton (Ithaca, NY).

This differs from the typical variety by the broader marginal area of rectangular epidermal cells on the ventral (abaxial) surface of the leaf, and by the slightly lower stature in development of papillae.  The Florida yew is exceptional in occurring at relatively low elevations in seasonally hot humid forests and also having leaves with a papillose midrib and angular (isodiametric) cells.  These features are usually found in yew species that grow above 2000 m.  At lower elevations, as seen for yews in  China, yew leaves generally lack papillae on the abaxial midrib, and their epidermal cells in T-section appear more rounded elliptical than angular.  

The Florida yew is clearly related to the Mesoamerican yew.  It may have found refuge at lower elevations along the Gulf as glaciers advanced during the Pleistocene. Compared to a wide range in differences seen in Asian yew, the Florida yew evidently has not had much time to differentiate itself from the Mesoamerican yew.

 

Taxus globosa var. globosa, showing marginal cells on abaxial leaf surface and marginal papillae, magnified 250x	Taxus globosa var. floridana, showing stomata band with 7 rows of stomata, and midrib and marginal cells in mid region of leaf on abaxial surface, magnified 250–400x
Taxus globosa var. floridana, lectotype (PH) and handwriting of Thomas Nuttall from diary  (J. E. Graustein.  1951.  Nuttall's travels into the old Northwest, an unpublished 1810 diary.  Chronica Botanica Co., Waltham, MA).	Taxus globosa var. floridana, specimen at K labeled T. croomii
Taxus globosa var. floridana, specimen at K labeled T. croomii from Bentham Herbarium	Taxus globosa var. floridana, specimen at K labeled T. floridana collected by Chapman
	Taxus globosa var. floridana, specimen at PH labeled T. croomii Chapman

Anderson, E. D. and J. D. Owens. 1999. Megagametophyte development, fertilization and cytoplasmic inheritance in Taxus brevifolia. Int. J. Plant Sci. 160: 459–469.

Baye, P. 2004. Comments. THP-1-04-059-SON Martin (Sleepy Hollow) Timber Conversion Permit/Timber/Harvest Plan (Brushy Ridge, Annapolis, Sonoma County, California), to the California Dept. Forestry, 15pp.. www.gualalariver.org/vineyards/martin-baye.pdf.

Clement-Westerhof, J. A. 1988. Morphology and phylogeny of Paleozoic conifers.  Pp. 298–337 in C. S. Beck (ed.), Origina and evolution of gymnosperms. Columbia Univ. Press, New York.

Foster, A. S. & E. M. Gifford, Jr. 1974. Comparative morphology of vascular plants.  2nd ed.  W. H. Freeman and Company, San Francisco.

Harris, T. M. 1976. The Mesozoic gymnosperms. Rev. Paleobot. and Palyn. 21: 119–134.

Judd, W.S., Campbell, S., Kellogg, E.A., Stevens, P. F. & M. J. Donoghue.  2002.  Plant Systematics.  A phylogenetic approach.  Sinauer Associates, Inc., Sunderland, MA.

Robertson, A. 1907. Taxoideae:  A phylogenetic study.  The New Phytologist VI: 92–102.