Circumscription and lectotypification of Hedychium villosum and its variety H. villosum var. tenuiflorum (Zingiberaceae)

Abstract The nomenclatural confusion between the Indian gingers Hedychium villosum Wallich and its variety Hedychium villosum var. tenuiflorum (Wall. ex Voigt) Wall. ex Baker is discussed. Both taxa are lectotypified in order to stabilize the names and contribute towards a resolution of their confusing nomenclatural past. Both taxa are described in detail to aid identification.


Introduction
Zingiberaceae are one among the ten largest monocotyledonous families in India, represented by 20 genera and about 200 species, and are mainly concentrated in the Northeastern, Peninsular and Andaman and Nicobar regions. Hedychium, commonly called "ginger lily" or "butterfl y lily" produces one of the most beautiful and aromatic fl owers in the family Zingiberaceae. Th e genus Hedychium was established by Koenig (1783) with H. coronarium J.Koenig as the only species, based on Rumphius's (1747) illustration. Hedychium includes about 80 species with highest species diversity in the eastern Himalaya region to South China and Southeast Asia. About half the species occur in the Indochinese region (Sirirugsa and Larsen 1995). Hedychium is the largest genus of Zingiberaceae in India, with about 43 taxa (Jain and Prakash 1995), mostly restricted to Northeastern India.
Whilst revising the species of Hedychium in India, nomenclatural and circumscription problems were encountered in relation to two plant names coined by Nathaniel Wallich, H. villosum Wall. and its variety, H. villosum var. tenuifl orum (Wall. ex Voigt) Wall. ex Baker. Th e plants of these two taxa are very peculiar in their sagittate anthers and winter fl owering ( December to April), whereas in other Indian species of Hedychium the anthers are oblong and plants fl ower during the monsoon season (June through October). For a long time, H. villosum was identifi ed as H. villosum var. tenuifl orum or vice-versa or the two taxa were even lumped as a single highly variable species. As a result of extensive studies of the literature, herbarium and live specimens, we are now able to identify and delimit these two taxa, and recognize them as distinct. Th e variety diff ers in its shorter infl orescence, larger white fl owers and larger anthers (see Table 1, Figs 1A, B). Confusion over taxon identity was in part caused by poor knowledge of the types of these names.

Nomenclatural discussion
Hedychium villosum was originally described by Wallich (in Roxburgh 1820) from a specimen sent to him by Mathew Robert Smith from the "mountains north-east of Bengal". He described it as a native of northeast of Bengal, fl owers in rainy season and "Kattia Ram Rait" as its local name (Khasee language). In describing the species he compared it with a live specimen of H. gracilie Roxb. and diff erentiated in its larger plant size, length and villosity of infl orescence, copious fascicled fl owers, and corolla lobes of equal length. From Wallich's description, it is evident that this is a species with a cylindrical infl orescence of 25-30.5 cm long, with small, pale yellow fl owers (corolla tube c. 3.8 cm long, bracts much shorter than tube, calyx somewhat shorter than tube, fi lament as long as tube and scarlet). At Kew we located a colour plate drawn for Wallich that closely resembles the description of H. villosum Wall. Wallich also incidentally mentioned "var. tenuifl orum" under this taxon in his catalogue (Numer. List. 1832) for Wallich 6545C,  (Wallich 1853(Wallich , 1855. Indeed, it is evident from Wallich's own statement ("they diff er not even as varieties I believe"), that he preferred not to recognize his "var. tenuifl orum" as a distinct taxon. Wallich (1832) cited it as "6545-C: Hedychium villosum Wall. ? var. tenuifl orum Wall. HBC" (Hortus Botanicus Calcutta) "e Sillet" (from Sylhet). A few years later, Roscoe (1827) while treating 'Monandrian Plants of the order Scitamineae', accepted Wallich's broad view and did not recognise any infraspecifi c taxa under H. villosum. It is evident from his observation, "the chief diversity we have observed between our plant and that described in the 'Flora Indica', is in the downy margin and mid-rib of the leaves, and in the colour of the fl owers, which in the 'Flora Indica' are described as of a pale yellow colour, whilst in ours they are a pure white", that he had not recognized any taxa at intraspecifi c level probably for the reason that the sagittate anther peculiar to H. villosum is also present in the other element (Wallich's "tenuifl orum"). He stated further that, "Dr. Wallich has also communicated another species under the name Hedychium tenuifl orum, which resembles the present, as well in the deep-lobed, undulated lip, as in the small sagittate anther..…". Interestingly, the description and fi gure provided by Roscoe (1827, t. 54) under the name H. villosum are an excellent match for 'var. tenuifl orum' (sensu the specimen Wallich 6545C), not for H. villosum (sensu Roxburgh 1820).
Voigt (1845) (Voigt 1845). We feel that these two taxa are distinct but that they only warrant recognition at varietal, rather than specifi c rank. Hence, we recognize two taxa: H. villosum var. villosum as the nominal species and a distinct variety, H. villosum var. tenuifl orum.

Hedychium villosum
Distribution. Northeastern India, Bangladesh, northern Myanmar, southern Yunnan and Guangxi provinces of China, Th ailand, Vietnam, and Malaysia from 600 to 1800 m (Wu and Larsen 2000).  cies by the nature of infl orescence, less than 1 mm long anthers, narrower leaves and smaller fl owers. As with Baker (1892), Schumann's description of the taxon (except "corolla prob. albae") -"Folia omnia summa ipsa sessilia stricte lanceolata longissime attenuato-acuminata et rostrata acutissima basi angustata utrinque glabra ....., Spica 25-30 cm longa anguste cylindrical ....., anther vix 1 mm longa" perfectly agrees with what is currently understood as H. villosum Wall. Inspection of specimens cited, such as Hooker & Th omson s.n. (K!), Griffi th 5661 (K!) and Prain 43 (CAL!) supports our contention that this is what Schumann did. Schumann (1904) also cited a specimen from Silhet, Wallich 6545A both for his H. tenuifl orum and H. villosum. Although he mentioned Robert Smith's introduced material in 1815 under H. villosum, his description does not agree perfectly with Wallich's protologue. Th ere are two sheets in the microfi che of the Wallich Catalogue bearing catalogue number 6545A. Of the two sheets labelled as Wall. Cat. n. 6545A, the fi rst sheet contains two specimens of H. villosum var. tenuifl orum and second one is a composite of two taxa, the top left material corresponding to H. villosum and the other two specimens matching what we currently recognize as H. villosum var. tenuifl orum. Judging from the description and materials cited by Schumann (1904), it appears that he treated H. villosum in the sense of Roscoe (1827), not in the sense of Wallich.