Romulea

Manning, John C. & Goldblatt, Peter, 2001, the Arabian Peninsula and Socotra including new species, biological notes, and a new infrageneric classification, Adansonia (3) 23 (1), pp. 59-108 : 64

Relationships of Romulea

Romulea is usually believed to be most closely allied to Syringodea because of their comparably reduced habit (all Syringodea species are acaulescent), branched inflorescences with actinomorphic flowers solitary on the branches and similar corms with woody tunics and a crescent-shaped basal ridge. There are, however, several fundamental differences between the two. The leaves of Syringodea are bifacial and usually channeled (or sometimes terete) and lack the grooves characteristic of Romulea ; the outer bracts in Syringodea are tubular below through fusion of the lower margins; the style branches in Syringodea are simple, unlike the divided style branches in Romulea ; and the pollen in Syringodea , like that of Crocus but unlike other genera in Crocoideae , is inaperturate. In the absence of evidence to the contrary both Romulea and Syringodea are presumed to be allied to the genera Hesperantha and Geissorhiza in a clade defined by the actinomorphic flower (now believed to be derived within Crocoideae ) and the asymmetric corm with woody tunics. If this proposed phylogeny is correct, the grooved leaf, inner bracts with membranous margins and divided style branches are synapomorphies for Romulea . The evidence for a relationship between Syringodea and Romulea , however, is weak at best, for a subterranean stem is not uncommon in the Iridaceae and an inflorescence with flowers solitary on the branches is a feature of most acaulescent Crocoideae , including Duthieastrum , Ixia acaulis and Crocus . Moreover, chloroplast DNA sequence data do not strongly associate the two genera nor, in fact, point to any particular relationships for Romulea ( REEVES et al. 2000) .