Arundinoideae. –This subfamily has presented problems to grass taxonomists for many years (Clayton & Renvoize 1986; Conert 1987; Ellis 1987; Hsiao et al. 1998; Kellogg & Campbell 1987; Renvoize 1981, Watson et al. 1985) as it has in the past being regarded primarily as repository for many genera that could not be readily accommodated in any of the other subfamilies (Renvoize 1981; Conert 1987). The study of Renvoize, based on characters from spikelet morphology and leaf anatomy separated out a core group from a peripheral group. On the basis of phylogenetic studies, based heavily on molecular data (Barker 1995: Linder 1997; Hsiao et al 1998; Barker 1999) the core group of Renvoize has been subdivided into the Danthonioideae (including the Cortiderieae) and Arundinoideae s.s., whereas the peripheral genera now almost all have been placed in other subfamilies. There is still no morphological synapomorphy for the subfamily as circumscribed in the molecular based phylogeny of GPWG (2001), where it is placed in the PACCAD clade between the Danthonioideae and the Chloridoideae. Unlike most other subfamilies where there has been recent phylogenetic investigation, the research focussing wholly or partially on the phylogeny of the Arundinoideae is more than 10 years old (Barker et al., 1995; Barker 1999; Hsaio et al. 1998; Linder et al. 1997) and the boundaries of both some genera and the subfamily as a whole are still not finalised. After the exclusion of genera now considered to be danthonioid or centothecoid, the 20 genera presently considered as belonging to the Arundinoideae are Alloechaete (Rendle) C.E.Hubb. (6 species from Africa), Amphipogon R.Br. ((9 species from Australia), Arundo L. (3 species from temperate Asia or cosmopolitan – A. donax), Centropodia Reichenb. (4 species from Africa and Asia), Crinipes Hochst. (2 species from N and E Africa), Danthonidium (1 species from India), Dichaetaria Nees (1 species from India), Dregeochloa Conert (2 species from the Cape and Namibia), Elytrophorus P. Beauv. (2 species from the Old World Tropics), Hackonechloa Makino (1 species from Japan), Leptagrostis C.E.Hubb. (1 species from Ethiopia), Molinia Schrank (2 species from North Africa and Japan), Monachather (1 species from Australia), Nematopoa C.E.Hubb. (1 species from Zambia and Zimbabwe), Phaenanthoecium C.E.Hubb. (1 species from Ethiopia and Sudan), Phragmites Adans. (4 species worldwide, although regarded as monotypic in North America (Barkworth et al. 2003), Piptophyllum C.E.Hubb. (1 species from Angola), Styppeoichloa de Winter (3 species from Southern Africa and Magadgascar) and Zenkeria Trin. (5 species from India). These genera are in need of re-analysis including representatives from all these genera and using the latest molecular techniques. On the basis of previous research some of the genera from the above list have been assigned to a crinipoid clade (Crinipes, Leptagrostis, Nematopoa, Piptophyllum, Styppeoichloa, Zenkeria) on the basis of morphological characters (Linder et al. 1997), but this needs substantiating using molecular methods. All the other genera (Amphipogon, Arundo, Centropodia, Danthonidium, Dichaetaria, Dregeochloa, Elytrophorus, Hackonechloa, Molinia, Monachather, Phaenanthoecium, Phragmites) are presumably members of a group known as ‘Arundineae sensu stricto’ (Linder et
al. 1997), but the relationships between them are little understood.
Ar Arundineae Am Amphipogoneae
Ar Eragrostis walteri