Under & Verboom, Generic limits in the Rytidosperma complex 
599 
Ainphipogoii and Diplopogon, which were included in the Danthonieae by Watson and 
Dallwitz (1992), are probably more distantly related, and may have closer relationships 
to other lineages. The exclusion of these taxa is based partially on the absence of the 
danthonioid synapomorphies, but in most cases the morphological evidence for the 
alternative placings has not yet been investigated, and current opinions on the affinities 
of these taxa are based on plastid genome sequences (rhcL and rpoC). 
Momchather Steudel was separated from Dantliouia s.l. by Blake (1972) on the basis 
of the broadly turbinate, indurated lemma and the subquadrate caryopsis with the 
small elliptical hilum, as well as the woolly plant base. In addition, Watson and 
Dallwitz (1992) documented the presence of a lemma with a germination flap, and 
the structure of the lemma lobes is quite unlike any found in the danthonioids (pers. 
obs.). Morphologically the genus is quite distinct, and has been generally accepted 
(Jacobs 1994, Walsh 1995), but its relationships have remained obscure. Molecular 
data (Barker 1995) suggest a relationship to the Arundineae s.s. rather than to the 
Danthonieae, but its high ploidy level and bizarre morphology suggest that a more 
detailed analysis might be necessary to assess congruence of the plastid and nuclear 
genomes. As the molecular data (Barker et al. 1995) and morphological data 
(unpublished) indicate a rather distant relationship to the Danthonieae, the genus 
was not included in the present analysis. 
It is clear that there is as yet no consensus as to what constitutes the best generic 
classification of the Australasian danthonioid grasses. This paper addresses the 
delimitation problems between Rytidosperma, Erythranthera, Pyrranthera, Monostachya 
and Danthonia, and constitutes an attempt to delimit monophyletic genera within 
this complex. 
Methodology 
Almost all species assigned to the danthonioid clade and related groups were studied 
morphologically and anatomically from herbarium material and, where available, 
from live plants. The descriptive data are being assembled in a DELTA data-base 
(Dallwitz 1980, Dallwitz & Paine 1986). Selected species were studied embryologically 
(see Verboom et al. 1994). Further information, especially cytological information, 
was taken from the literature (for Australian material. Brock & Brown 1961). The 
Australian species were studied from a large number of specimens, in preparation 
for the Flora of Australia account. 
Spikelet, floret and caryopsis morphology were recorded both from whole-mounts 
of dissected spikelets in glycerine, and from dry dissections; measurements were 
taken with an eyepiece graticule precise to the nearest 0.1 mm, and drawings were 
prepared by camera lucida. Continuous characters were used as ratios to prevent size 
variation having an unduly large effect. The ratios were divided into states at 
convenient intervals to reflect the range of the character variation: the character 
variation was not assessed for 'real' intervals in the variation range. Flowever, as 
few taxa were polymorphic for the continuous characters, it is assumed that these 
intervals may be informative on relationships between taxa. 
Leaf anatomical preparations were made from the midportions of the lower leaves 
fixed in FAA in the field, and stored in 70% EtOH. Where freshly fixed material was 
not available, herbarium material was rehydrated in boiling, soapy water. Transverse 
sections were hand-cut, while adaxial and abaxial epidermal scrapes were prepared 
by standard techniques. Preparations were stained in a combined Alcian Blue - 
Safranin stain (Tolivia & Tolivia 1987), dehydrated through an alcohol series, and 
mounted in Canada Balsam. 
