The taxonomic position of Tricoryne R. Br. has been investigated employing quantitative taxonomic methods involving 27 mainly Australian endemic genera (28 OTUs) and 60 of their attributes.
Living material was studied, unless unavailable, when data were taken from preserved material or the literature. The taxonomic role of several anatomical and morphological characters in particular, not previously widely used in the supra-generic classification of the lilies, have been shown to be significant. Thus, the occurrence of root endodermis thickening and many pollen characters obtained from scanning electron microscopic studies especially symmetry, shape, aperture type and exine sculpturing are important for defining the clade which corresponds with Phormiaceae.
The work of Dahlgren et al. (1985) was adopted as a basis of classification and as a source of information on relationships against which the results obtained were compared.
For the phenetic analyses, Euclidean distance was used as a measure of dissimilarity and the intertaxa distances were subjected to two clustering strategies, Group Average and Incremental Sum of Squares. The traditional Phormiaceae assembled as a single group while Anthericaceae sensu lato was fragmented. Furthermore, the two genera of Asphodelaceae included were isolated from Phormiaceae and were located near to members of the fragmented Anthericaceae.
A wide variety of cladistic analyses employing parsimony and character compatibility algorithms were undertaken. The parsimony methods, with character reversals permitted, generated structured
cladograms with generic relationships reflecting closely one of the phenetic analyses. With a few minor differences in the positions of some genera, the cladogram arising from the parsimony method with character...............