Of the approximately 380 families of angiosperms, representatives of only 10 are known to form symbiotic associations with nitrogen-fixing bacteria in root nodules. The

Analysis of sequences of the chloroplast gene rbcL for 76 taxa of Podocarpaceae (representing all genera except Parasitaxus) and five species of Phyllocladaceae were undertaken with respect to their relationships to each other and to 28 coniferalean outgroup taxa from seven families. The results indicate that Podocarpaceae are polyphyletic unless expanded to include Phyllocladaceae. Within Podocarpaceae, Sundacarpus is placed in a clade with Prumnopitys, and Falcatifolium is paraphyletic as a basal grade to Dacrydium. Phyllocladus is in an unresloved clade with Halocarpus, Manoao/Lagarostrobos and Prumnopitys/Sundacarpus. The separation of Afrocarpus from Podocarpus and its placement instead as sister to Nageia and Retrophyllum is supported. Podocarpus s. str. is monophyletic, with both subgenera identified, albeit poorly supported. The analysis placed Lepidothamnus and Saxegothaea in an unresolved basal polytomy within the family. There were no clear outgroup relationships with the family. These results differ from the morphological clades found by Kelch (1997), and disagree strongly with his 18S-sequence-based phylogeny (Kelch 1998). However, jackknife support values indicate that although the genera are well supported, relationships both within and between them are not, suggesting that intergeneric relationships in the family require further study. There is also some congruence between our results and those of the gymnosperm 18S study by Chaw et al. (1997), although their study included only three Podocarpaceae and one Phyllocladaceae species.

Sequences of rbcL for 23 species of Nothofagus and three of Fagus have been determined and analysed to form phylogenetic trees. The two genera are well separated. The species of Nothofagus separate into lineages which correspond exactly with the subgenera recently defined on morphological grounds. The rate of evolution of the four subgenera is shown to be statistically the same and, using a reference date from palaeobotany, is found to be one nucleotide change in 6 Ma. This rate is used to derive the ages of the common ancestors of species in subgenera and it is tentatively concluded that intercontinental dispersal was possible in the early stages of the evolution of the genus.

No abstract

No abstract