紫菀属 zi wan shu

Authors: Yilin Chen, Luc Brouillet & John C. Semple

Aster diplostephioides Credit: Harvard University Herbaria

Herbs, perennial, rarely annual or biennial, subshrubs or shrubs, generally rhizomatous, sometimes somewhat woody. Stems erect, sometimes decumbent to ascending, or procumbent. Leaves alternate, rarely opposite, sessile or petiolate, obovate to elliptic-ovate, ovate, or lanceolate, margin entire, serrate, or coarsely dentate, sometimes pinnatifid. Capitula solitary or in corymbiform or sometimes paniculiform synflorescences, radiate, rarely discoid. Involucres hemispheric, campanulate, or obconic; phyllaries 2-5-seriate, imbricate or equal, linear-lanceolate, herbaceous or membranous, margin scarious (at least inner), apex sometimes purplish. Receptacles flat or convex to conic, alveolate, sometimes lacerate, epaleate. Ray florets female, fertile or rarely sterile, 1- or 2-seriate, rarely absent, white, pink, purple, or blue, apex inconspicuously 2- or 3-denticulate; disk florets numerous, bisexual, fertile, yellow or purple-brown, limb campanulate, 5-lobed, lobes equal or unequal (1 lobe deeper and corolla zygomorphic). Anther base obtuse; style branch tip lanceolate or triangular. Achenes oblong, obovoid, or oblanceoloid, compressed or subconvex, rarely 3- or 4-angled, margin 2(-6)-ribbed, glabrous or strigose, eglandular or glandular. Pappus 1-4-seriate, persistent or rarely caducous, sometimes absent, white, brownish, or reddish, outer series of short slender bristles or scales, inner series of numerous subequal, barbellate or barbellulate bristles, apex acute or sometimes innermost clavate, sometimes of free short bristles or scales or connate scales and inner bristles absent, sometimes ray pappus absent and disk present.

About 152 species: Asia, Europe, North America; 123 species (82 endemic) in China.

The circumscription of Aster adopted here corresponds to that used in recent Asian floras (Ito & Soejima in Iwatsuki et al., Fl. Japan 3b: 59-73. 1995; Soejima & C. I Peng, Fl. Taiwan, ed. 2, 4: 848-868. 1998), which excludes North American segregates such as Doellingeria Nees, s.s., Eurybia, and Symphyotrichum (Nesom, Phytologia 77: 141-297. 1994; Semple & Chmielewski, Fl. N. Amer. 20: 43-46. 2006; Brouillet, Fl. N. Amer. 20: 365-382. 2006; Brouillet et al., Fl. N. Amer. 20: 465-539. 2006), while including genera such as the Asian members of Doellingeria (sensu Nesom, loc. cit.) (A. sect. Teretiachaenium), Kalimeris (A. sect. Asteromoea), Heteropappus (A. sect. Pseudocalimeris), and Miyamayomena and Rhynchospermum (both included within A. sect. Aster). Such delimitation is supported by recent molecular phylogenetic analyses of tribe Astereae, summarized in Brouillet et al. (in Funk et al., Syst. Evol. Biogeogr. Compositae, 589-629. 2009). Other genera that might probably be considered part of Aster are Sheareria, which Gao et al. (Taxon 58: 769-780. 2009) placed close to Kalimeris and Heteropappus but those were placed within the Aster complex as described here in a more recent analysis of Iranian Astereae involving more members of the genus (F. Jafari, pers. comm.), and Rhinactinidia, a possible sister of A. sect. Pseudocalimeris (F. Jafari, pers. comm.). This last relationship would be supported by the fact that these two groups share unequal disk corolla lobes, a rare feature in the tribe. Other genera may be involved here, but this is still unclear. The current concept excludes African Aster species, which belong to a distinct, African clade (Brouillet et al., loc. cit. 2009)

An alternate classification that would retain all segregate genera as distinct would recognize Heteropappus, Kalimeris, Rhinactinidia (though one could defend including them in Heteropappus), Sheareria, and Aster s.s. (still including Miyamayomena and Rhynchospermum) and would impose the creation of new genera for A. sect. Teretiachaenium and A. sect. Ageratoides (unless one prefers a strongly paraphyletic, not to say polyphyletic, Aster).

Nesom (loc. cit.) underlined the distinction of the shrubby species of Aster ser. Albescentes Y. Ling, noting their isolation among Old World asters. He concluded that the segregation of the series at generic rank would be justifiable. Two of us (Brouillet and Semple) agree with this hypothesis, but, currently, phylogenetic data are insufficient to place this group within the phylogeny of the tribe Astereae. Therefore, we treat it here as an incertae sedis group within Aster for the time being. One of us (Chen) considers that A. ser. Albescentes should be recognized as a section; however, the combination at that rank is not formally proposed here because not all the present authors agree.

Aster filipes J. Q. Fu (Bull. Bot. Res., Harbin 3(1): 118. 1983), described from Gansu, and A. langaoensis J. Q. Fu (loc. cit.), described from Shaanxi, could not be treated here because no material was seen by the present authors.