33. Rubia tibetica J. D. Hooker, Fl. Brit. India. 3: 204. 1881.
西藏茜草 xi zang qian cao
Herbs, perennial, erect, or subshrubs, forming loose cushions from a massive woody rootstock; stems to 0.3 m, quadrangular, glabrous or mostly with ± dense hirtellous indumentum, on angles shortly aculeolate and hispidulous with partly hooked trichomes, lower nodes usually shortly sheathed with membranous bases of old leaves. Leaves opposite and with interpetiolar and ± leaflike stipules in whorls of 4(-6), sessile or subsessile; blade drying leathery, broadly to narrowly ovate, elliptic, elliptic-oblong, or lanceolate, 1-3 × 0.3-1.5 cm, both surfaces hirtellous to ± glabrous, base acute to obtuse, margins retrorsely aculeolate, toward acute and often cuspidate apex usually antrorsely aculeolate; principal vein 1, sometimes with 2 weak lateral veins. Inflorescences leafy and bracteose, with axillary and terminal, 1- to few-flowered cymes; axes mostly glabrous; pedicels (2-)4-6(-14) mm. Ovary 0.8-1.2 mm, sometimes hirtellous. Corolla yellow, rotate, 5-8 mm in diam., outside sometimes scaberulous; fused base ca. 0.5 mm; lobes lanceolate or lanceolate-ovate, 2-2.5 mm, acuminate. Mericarp berry 3-4 mm in diam. Fl. Jun, fr. Aug.
In gravel at river bottoms; [1700-]3600[-4400] m. Xinjiang, Xizang [Afghanistan, India (Punjab), Kashmir, Kyrgyzstan, Pakistan, Tajikistan].
Rubia tibetica is a relatively widespread SW to C Asiatic and typically montane to alpine pioneer species. It was illustrated by Deb and Malick (Bull. Bot. Surv. India 10(1): 4, f. 3. 1968), who lectotypified its name with a specimen from "Tibet" (i.e., Xinjiang). The species is notable for its shoot morphology: the well-developed vegetative leaves of lower stem nodes often appear in a 4-verticillate arrangement with ± leaflike stipules, whereas the basalmost first leaves produced by the stems and also the leaves of reproductive nodes are generally paired and exhibit typical interpetiolar stipules.
Deb and Malick (loc. cit.: 4-5) described Rubia aitchisonii Deb & Malick from Bagdis, Afghanistan, and separated it by: "Lamina ovate, sub-orbicular, elliptic-ovate or lanceolate, 2-6 in a whorl" in R. tibetica vs. "Lamina elliptic-lanceolate, 2 opposite" in R. aitchisonii. Ehrendorfer and Schönbeck-Temesy (Fl. Iranica 176: 67. 2005) noted that the only locality of R. aitchisonii lies within the area of R. tibetica and that the suspected specific differences fall within the morphological variability of R. tibetica. Thus, future studies may show that R. aitchisonii is better synonymized under R. tibetica.
Rubia tibetica was placed by Pojarkova (Fl. URSS 23: 401-404. 1958) into R. ser. Tibeticae Pojarkova in R. sect. Campylanthera together with two other alpine, C Asiatic species: R. regelii Pojarkova and R. komarovii Pojarkova. They differ from R. tibetica by leaves and leaflike stipules in whorls of up to 6 but have not been recorded yet from N China. Rubia garrettii Craib from Thailand, also suspected to be a member of this group, certainly does not belong here (see Puff, Fl. Thailand: Rubiaceae; http://homepage.univie.ac.at/christian.puff/FTH-RUB/FTH-RUB_HOME.htm; accessed on 5 Oct 2010). Whereas the R. tibetica group is provisionally included in R. sect. Rubia, R. garrettii obviously belongs to R. sect. Oligoneura.