Systematics of Scleranthus (Caryophyllaceae)
Scleranthus is a genus of about 12 species of herbaceous flowering plants or small shrubs with a disjunct Eurasian/Australasian distribution. Monophyly of the genus is supported by the close similarity of gynoecial development of all species and consistent with nuclear ITS DNA sequence analysis. Traditionally the genus had been divided into two sections, section Scleranthus and section Mniarum. Section Mniarum is exclusively Australasian while section Scleranthus has been circumscribed to contain exclusively European species or a combination of European and Australasian species. Pollen and floral characters align the species into Australasian and Eurasian groups also supported by nuclear ITS DNA sequence analysis. Section Scleranthus as more broadly defined (i.e., sensu West and Garnock-Jones, 1986) is therefore at least paraphyletic or at worst polypyhyletic. Phylogenetic reconstructions based on morphological characters differ from those based on ITS sequences in supporting different relationships within the Australasian species of Scleranthus. Hybridisation and introgression within the genus are discussed and suggested as the cause of discordance between morphology and DNA sequence based trees. Low sequence divergence among Scleranthus ITS sequences suggests that the European and Australasian clades within the genus diverged within the last l0 million years. Biogeographic implications of these dating and competing hypotheses explaining the disjunct North-South distribution of the genus are discussed. Nuclear ITS and chloroplast ndhF DNA sequences both suggest that Scleranthus belongs to a clade within the family Caryophyllaceae consisting of members of subfamilies Alsinoideae and Caryophylloideae. Phylogenetic relationships between genera belonging to the three subfamilies of Caryophyllaceae (Alsinoideae, Caryophyloideae, and Paronychioideae) are addressed in this thesis through ndhF sequence analysis, which provides no support for the monophyly of traditionally recognised groups. Morphological character data sets are likely to always encompass multiple incongruent data partitions (sensu Bull et al. 1993). It may therefore be appropriate to combine data from DNA sequence and morphology for parsimony analysis even where the two are significantly incongruent.