3. Amentotaxus Pilger, Bot. Jahrb. Syst. 54: 41. 1916.
穗花杉属 sui hua shan shu
Shrubs or small trees dioecious; branchlets opposite, base with bud scales not persistent; winter buds tetragonal-ovate, acute, glossy, scales, decussate, in 3-5 whorles of 4, ridged adaxially. Leaves decussate, but brought into a single plane by twisting of petioles; blade straight or slightly falcate, usually more than 5 mm wide, adaxial surface mottled when fresh, rarely smooth, rugose or ± so when dry, resin canal present below sheath of vascular bundle, sclereids present, rarely absent, base decurrent, margin slightly downcurved. Pollen cones aggregated into (1 or)2-6(-10) long, slender, compound racemes or spikes arising from bract axil near apex of branches; individual cones opposite, sessile or subsessile, ellipsoid or subglobose; microsporophylls numerous, ± shield-shaped; pollen sacs 3-8, arranged radially, or adaxially and abaxially. Seed-bearing structures compressed-tetragonal or flattened abaxially, basal part with 6-10 pairs of decussate bracts arranged in 4 rows each of 3-5 bracts; ovule 1, sessile, erect. Seed ripening in 1st year, long pedunculate, ellipsoid or obovoid-ellipsoid, enclosed except for apex in a saclike aril which is bright red or reddish yellow when ripe; bracts persistent at base.
Five or six species: China, Vietnam; three species (one endemic) in China.
Amentotaxus has been placed in its own tribe (Amentotaxeae W. C. Cheng & C. D. Chu) or family, Amentotaxaceae. Although Page (in Kubitzki, Fam. Gen. Vasc. Pl. 1: 299-302. 1990)
included it in the Cephalotaxaceae, molecular data indicate very strongly that it is the sister genus of Torreya, which is similar in the size and shape of its seed and in usually having
bilaterally symmetric clusters of pollen sacs.
Amentotaxus assamica D. K. Ferguson (Kew Bull. 40: 115. 1985) was described from SE Xizang, in temperate rainforests on steep, north-facing slopes, associated with species of
Magnolia, Quercus, Rhododendron, etc. It is apparently similar to A. argotaenia, but differs in having leaves without sclereids, the adaxial surface smooth or with only longitudinal
striations (due to shrinkage) when dry. Further collections are needed to establish whether it is truly distinct from A. argotaenia.
