Danthonioideae. --The subfamily Danthonioideae has only been recognised as being separate from the Arundinoideae for a relatively short time (GPWG, 2001). A historical perspective on the subfamily (Linder & Barker 2005) shows how the composition of the subfamily has progressed from its early traditional admixture with the arundinoids to the recognition of a separate subfamily on phylogenetic principles based on molecular (Barker et al. 1995) and anatomical (Verboom et al.1994) data.
As presently circumscribed (Linder et al, 2010) it constitutes a small, well-defined clade comprising 282 largely Southern Hemisphere grass species in 17 genera - Austroderia N.P. Barker & H.P.Linder, Capeochloa H.P.Linder & N.P.Barker, Chaetobromus Nees, Chionochloa Zotov, Cortaderia Stapf, Danthonia DC., Geochloa H.P.Linder & N.P.Barker, Chaemaerochloa H.P.Linder, Merxmuelleria Conert, Notochloë Domin, Pentameris P.Beauv., Plinthanthesis Steud., Pseudopentameris Conert, Rytidosperma Steud., Schismus P. Beauv, Tenaxia N.P. Barker & H.P.Linder and Tribolium Desv. With the exception of Danthonia (North America and Europe) and Schismus (Mediterranean Europe and South Africa) all these genera have their main representation in the Southern Hemisphere or extensions northwards in Africa and South America in the elevated montane zones. Other than Rytidosperma with species in New Guinea, Australia, New Zealand and South America all the other genera are more or less endemic to one of the southern continents or islands. Eight are endemic to Africa (Capeochloa, Chaetobromus, Geoochloa,
Merxmuelleria, Pentameris, Prionanthium, Pseudopentameris, Tribolium), one (Austroderia) is endemic to NewZealand and one (Chionochloa) has most species endemic there, two (Notochloë and Plinthanthesis) are endemic to Australia, one (Chimaerochloa) is endemic to New Guinea, and one (Cortaderia) endemic to South America. Two genera Monachather and Elytrophorus, recently assigned to the Danthonioideae (Linder in Mallett, 2005) are now not thought to belong in this subfamily (Linder pers. comm.), so in this account they have been assigned to the subfamily Arundinoideae, where they have been previously placed (Barker, 1997).
Prior to the new classification of the Danthonioideae, based on a well resolved molecular phylogeny of almost all the genera and supported by morphological input, previous cladistic work, based only on morphological characters, generated a rather different classification, particularly in the Australasian region. A cladistic analysis of 70 species of mainly Australasian species of danthonioid genera using morphological data (49 characters) by Linder and Verboom (1996) was the basis for a new classification of this group for this region at the time. They proposed a division of the Australasian representatives of Rytidosperma
Steud. (sens. lat.) into 3 smaller genera Austrodanthonia H.P. Linder, Notodanthonia Zotov and Rytidosperma sens. strict. on rather weak morphological evidence. There was no unique synapomorphy for Notodanthonia and those synapomorphies for Austrodanthonia (long callus) and Rytidosperma (small metaxylem vessels) were difficult to detect. Although Australian publications have taken up the generic splits (Sharp & Simon, 2002; Mallett, 2005), the grass volume for the Flora of New Zealand (Edgar & Connor, 2000) does not, only recognising Rytidosperma sens.lat. A similar view of this danthonioid group is followed at Kew as reflected in GrassBase and it now appears that this broad concept of Rytidosperma is more acceptable (Linder et al, 2010). This is corroborated by work on African members of the Rytidosperma clade (Verboom et al. 2006), which in addition to revealing a broad Rytidosperma, presents a topology in which the African genera of Karroochloa Conert & Túrpe, Schismus P. Beauv, and Tribolium Desv. are shown to be polyphyletic on the basis of plastid and nuclear sequences. This approach had been used to show that Cortaderia, as then circumscribed, was paraphyletic (Barker et al, 2003) and was extended to the phylogeny of all of the Danthonioideae (Barker et al., 2007) where the paraphyly of Cortaderia was presented again, as well that of Merxmuelleria.
H. Peter Linder, Marcelo Baeza, Nigel P. Barker, Chloe´ Galley, Aelys M. Humphreys, Kelvin M. Lloyd, David A. Orlovich, Michael D. Pirie, Bryan K. Simon, Neville Walsh, and G. Anthony Verboom (2010). Ann. Mo. Bot. Gard. 97: 320, fig 2.
Pirie - Monocots 4 (2008)