Evolutionary Studies in Asterids Emphasising Euasterids II

This thesis deals with evolutionary relationships within the asterids, a group of plants comprising about one-third of all flowering plants. Two new families are recognised: Pennantiaceae and Stemonuraceae. The woody Pennantia from New Zealand and Australia is the sole genus of Pennantiaceae. Stemonuraceae consist of a dozen woody genera with a pantropical distribution and a centre of diversity in South East Asia and the Malesian islands. They are characterised by long hairs on their stamens and/or fleshy appendages on their fruits. Both families were formerly included in Icacinaceae. While Pennantiaceae are unrelated to any of the former Icacinaceae and placed in the order Apiales, other former Icacinaceae genera are related to Cardiopteris, a twining herb from South East Asia and Malesia. The monogeneric family Cardiopteridaceae is enlarged as to include also these. Cardiopteridaceae and Stemonuraceae are sister groups and placed in Aquifoliales. The three other families of Aquifoliales are monogeneric and closely related. The Asian Helwingiaceae and the Central/South American Phyllonomaceae are suggested to be merged into Aquifoliaceae (hollies). The genera of Icacinaceae in the traditional sense not placed in any of the above families (all euasterids II) are members of early diverging lineages of the euasterids I and possibly included in the order Garryales. The three woody Australasian families Alseuosmiaceae, Argophyllaceae, and Phellinaceae are confirmed as members of Asterales, despite traditional placements not close to that order. They are, moreover, supported as each other’s closest relatives. The results are based mainly on parsimony analysis of DNA sequence data, but morphological studies have revealed characters in support for the molecularly based conclusions. The gene that has provided most new information is the chloroplast ndhF gene. The results are, however, drawn from combined analyses of sequences from one or several additional genes (atpB, matK, rbcL, 18S rDNA). The data have also been explored with Bayesian analysis, a statistical, model-based method that most recently has been developed for phylogeny reconstruction