Gnaphalium L.
Herbs, rarely woody at base. Lvs alternate, simple, entire. Capitula solitary or in small to large clusters; clusters usually capitate or forming spikes, rarely forming panicles. Involucral bracts in 2-several rows, imbricate, membranous, sometimes with coloured but not radiating lamina. Receptacle ± flat; scales 0. Outer florets ♀, filiform, usually many. Inner florets ⚥, tubular, few. Achenes all similar, ± terete, not angled or ribbed; pappus hairs in 1 row, narrow at apex.
Key
c. 150 spp., cosmopolitan. Native spp. 14, naturalised 8.
Gnaphalium has long been considered rather a heterogeneous assemblage of spp., and Drury (1970, op. cit.), in providing a fresh approach to its classification, recognised 3 main groups based mainly on the form of the involucral bracts (Fig. 26) - of these, the native spp. treated by him as anaphalioid cudweeds have been transferred to Anaphalis (Webb, in Connor and Edgar, op. cit.), while the sp. previously known as G. candidissimum is treated in Vellereophyton (Hilliard and Burtt, op. cit.); G. luteoalbum, classified by Drury as an achyroclinoid cudweed is treated within a broader concept of Pseudognaphalium Kirpiczn. (Hilliard and Burtt, op. cit.). Within the remaining group, the gnaphalioid cudweeds, Drury recognised 4 sections and these are sometimes treated at generic level. The indigenous spp. all fall within sect. Euchiton, the naturalised American spp. within sect. Gamochaeta, and G. uliginosum within sect. Gnaphalium.
Drury, D. G., New Zealand J. Bot. 9 : 157-185 (1971), revised the American spicate cudweeds naturalised in N.Z. He distinguished 8 spp., several of them known from only a few collections, and provided illustrations for all of them. All of these taxa are accepted here, but one only at varietal rank. Of these, G. coarctatum is by far the most common and widespread. Allan (1961) accepted 10 spp. of Gnaphalium indigenous to N.Z.; of these 4 are referred here to Anaphalis, and 1 to Pseudognaphalium. The remaining spp., all in sect. Euchiton, were revised by Drury, D. G., New Zealand J. Bot. 10 : 112-179 (1972), who accepted 11 spp. All of Drury's spp. are accepted here and in addition 2 entities he treated at an infraspecific level are accepted here as full spp. Plant habit, lf shape and hairiness, involucral bract length, and achene hairs are all important characters for identification and Drury provided clear illustrations for most spp. Drury also provided ⚥ and ♀ floret numbers which may help to discriminate spp. In spite of the excellent work of Drury, difficulties remain within sect. Euchiton in N.Z. At the end of his revision he gave descriptions and listed collections for 3 sets (" examples") of plants of uncertain status. Plants of his examples 1 and 2 are referred by A. P. Druce to G. delicatum and G. audax sens. strict. respectively, and plants corresponding to example 3 are now treated as a new sp., G. laterale. Other particular difficulties are discussed under G. involucratum and G. limosum.
Descriptions of the involucral bracts for both the naturalised spp. of sect. Gamochaeta and the indigenous spp. in sect. Euchiton are based on the largest, fully developed bracts of capitula at anthesis.