Tetraplasia Rehder.

Shrubs, sometimes with paired infrastipular or superaxillary spines; branches sometimes with complex sympodial growth with reduced internodes and prophylls (Damnacanthus indicus); roots at least sometimes moniliform (i.e., nodose or "node-like constricted"). Raphides present. Leaves opposite, apparently without domatia; stipules persistent and becoming hardened or sometimes falling by fragmentation, interpetiolar or shortly united around stem, generally triangular, acute or shortly bifid to multifid. Inflorescences pseudoaxillary, superaxillary, apparently terminal, and/or paired on short shoots giving an appearance of being axillary, 1-flowered or usually cymose to fasciculate and 2-4-flowered, subsessile to apparently shortly pedunculate (i.e., borne on a leafless short shoot), bracteate with bracts usually small and glandular-fimbriate. Flowers subsessile to pedicellate and often nodding, bisexual, monomorphic or distylous. Calyx limb cupular or campanulate, 4-lobed (or 5-lobed, D. henryi). Corolla white to yellow or pale purple, tubular-funnelform, often leathery, inside densely pubescent in throat to throughout; tube rarely fenestrate (D. henryi); lobes 4 (or 5, D. henryi), valvate in bud. Stamens 4, inserted in upper part of corolla tube, included or exserted; filaments short; anthers dorsifixed. Ovary 4-celled (or 2-celled, D. henryi), ovules 1 in each cell and attached near top of septum, campylotropous; stigmas 4 (or 2, D. henryi), linear, included or exserted. Fruit red, drupaceous, globose to ellipsoid or oblate, fleshy, with calyx limb persistent; pyrenes 4(or 2, D. henryi), each with 1 seed, plano-convex, subglobose, ellipsoid, or obtusely trigonous; seeds medium-sized, subglobose to plano-convex; endosperm corneous; embryo small; radicle hypogeous.

The morphology of Damnacanthus was reviewed in detail by Robbrecht et al. (Blumea 35: 307-345. 1991), who described its complex, sympodial growth pattern and variations found in different species. They considered the stems of Damnacanthus to be composed of sympodial units, with varying degrees of development of the individual parts in different species. The most characteristic and apparently complicated growth is found in D. indicus, in which each sympodial unit comprises a basal node bearing a pair of prophylls, similar to bud scales or reduced leaves; then, a developed internode; then, a node bearing a pair of normally developed foliage leaves, decussate in orientation to the prophylls; then, a node (without an intervening internode) bearing a pair of thorns, decussate in orientation to the leaves. Growth of the stem continues from one of the axillary buds of the foliage leaves, which gives the thorns the appearance of being stipular or superaxillary in position. The alternating prophylls and foliage leaves produce the characteristic heterophyllous growth of this genus. Species of Damnacanthus vary in the characteristic number of nodes in each sympodial unit and in the development (or not) of the thorns. Robbrecht et al. (loc. cit.) interpreted the characteristic "spines" of the genus as reduced shoot systems produced from the axillary buds subtending two undeveloped (and thus missing) leaves. They considered the characteristic 2- or 4-flowered inflorescences of Damnacanthus to be formed of one or two sympodial growth units, produced from one or both axils of a node bearing foliage leaves, with each of these units comprising three congested nodes, lacking separating internodes, with the basalmost nodes bearing bractlike scales and the terminal node producing a flower in each axil then stopping growth, thus comprising a 2-flowered cymule. They also noted that, although previous authors have described the ovules of Damnacanthus as amphitropous or pendulous, in fact the ovules are unique in the Rubiaceae in being campylotropous.

Damnacanthus is represented by at least three species in Japan (Fl. Japan 3a: 224-225. 1993), several of them apparently common and hybridizing, and its taxonomy has been rather intensively studied there and in Taiwan (e.g., Koidzumi, Acta Phytotax. Geobot. 3: 155-160. 1934), where it apparently has some medicinal use. Damnacanthus was revised for China by H. S. Lo (Acta Phytotax. Sin. 17(3): 104-109. 1979), then treated comprehensively by Y. Z. Ruan in FRPS (71(2): 167-176. 1999) in an essentially monographic work. Koidzumi (loc. cit.) recognized three sections within Damnacanthus and treated Tetraplasia as a separate genus, based largely on root characters, but these were not mentioned again until Y. Z. Ruan (loc. cit.: 169) recognized two sections in Chinese Damnacanthus, one of them under an unpublished name. The Chinese plants with spines were included in D. sect. Damnacanthus; the unarmed plants were separated by Koidzumi in Tetraplasia and were included by Ruan in his second, unnamed section.

Naiki and Nagamasu (J. Pl. Res. 116: 105-113. 2003; Amer. J. Bot. 91: 664-671. 2004) surveyed the breeding biology of several Japanese and Chinese species of Damnacanthus and found variation in breeding system among species, discovered a correlation between ploidy with breeding system but not leaf size, and reported distyly and dimorphic pollen in this genus.

About 13 species: China, N India, Japan, Korea, Laos, Myanmar, Vietnam; 11 species (six endemic) in China.

(Authors: Chen Tao (陈涛); Charlotte M. Taylor)