Systematics of the Relhaniinae (Asteraceae-Gnaphalieae) in southern Africa : geography and evolution in an endemic Cape plant lineage

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title

University of Cape Town

The Greater Cape Floristic Region (GCFR) houses a flora unique for its diversity and high endemicity. A large amount of the diversity is housed in just a few lineages, presumed to have radiated in the region. For many of these lineages there is no robust phylogenetic hypothesis of relationships, and few Cape plants have been examined for the spatial distribution of their population genetic variation. Such studies are especially relevant for the Cape where high rates of species diversification and the ongoing maintenance of species proliferation is hypothesised. Subtribe Relhaniinae of the daisy tribe Gnaphalieae is one such little-studied lineage. The taxonomic circumscription of this subtribe, the biogeography of its early diversification and its relationships to other members of the Gnaphalieae are elucidated by means of a dated phylogenetic hypothesis. Molecular DNA sequence data from both chloroplast and nuclear genomes are used to reconstruct evolutionary history using parsimony and Bayesian tools for phylogeny estimation. The subtribe is found to be nonmonophyletic, and three independent Cape-centred clades are identified. These constitute the earliest-diverging lineages in the Gnaphalieae. Ancestral areas are reconstructed onto the tree using maximum likelihood and indicate a Cape origin and diversification for the Gnaphalieae, with subsequent, multiple dispersals out of southern Africa. A relaxed Bayesian clock is used to estimate dates for important events, calibrated using a combination of secondary age estimates, fossil pollen and geological events. The single inferred dispersal to Australasia, which may be responsible for the large diversity of the tribe there, is compared with other dated trans-Indian Ocean disjunctions, and results in a hypothesis of direct Miocene longdistance trans-oceanic dispersal facilitated by the West Wind Drift. Important directions for future biogeographic and systematic studies on the Gnaphalieae are identified. One of the Cape clades identified in the above study, the Stoebe clade, is investigated using the same molecular markers and an additional chloroplast DNA region, in order to reconstruct relationships amongst the species and test generic circumscription. One genus, Amphiglossa, is found to be non-monophyletic and a recommendation is made to resurrect the genus Pterothrix in order to maintain monophyly. Relationships amongst subclades in the Stoebe clade can not be resolved, possibly due to a paucity of chloroplast DNA polymorphisms. Ancestral-state reconstruction using maximum likelihood indicates a high degree of homoplasy in all the macro-morphological characters used by previous workers to delimit the genera that fall within this polytomy (Disparago, Elytropappus and Stoebe). The evolution of these leaf, synflorescence and floral morphological characters is examined in order to shed light on the degree of homoplasy exhibited by each, and the patterns of character-state change across the tree. The analysis provides insight into patterns of morphological evolution, laying the groundwork for re-evaluation of characters previously thought to be homologous. The need for a rigorous, systematic anatomical study including all members of the clade is highlighted and some previously little-explored characters are identified that might provide greater future taxonomic insight. Due to a poor palaeoclimatic record, the impact of Pleistocene climatic fluctuations on the palaeodistributions of plant species in the Greater Cape Floristic Region is largely unknown. One of the members of the Stoebe clade, the common and widespread shrub Elytropappus rhinocerotis, is examined using population genetic tools in order to explore the spatial distribution of genetic variation and to infer the degree and pattern of responses (range shifts, population extinction) to recent (Quaternary) climatic changes. A molecular fingerprinting tool, inter-simple sequence repeat (ISSR) PCR is used to examine populations from across the range of E. rhinocerotis in order to compare patterns of population genetic diversity with the long-noted and strong spatial patterns of plant species diversity in the CFR. A large amount of genetic variation is detected, which is apportioned largely amongst individuals within populations rather than amongst populations or regions. This is to be expected for an outcrossing and well-dispersed plant species. However, there is significant spatial structure and a very uneven distribution of diversity across the geographic range of E. rhinocerotis. Although areas that have high species-diversity also generally exhibited high genetic diversity, the eastern margin of the range and the Kamiesberg highlands both have a far higher relative genetic diversity in E. rhinocerotis than would be predicted by their relative species diversity. This indicates that the processes producing greater leves of species diversity in some parts of the CFR are different than the processes responsible for high genetic variation within E. rhinocerotis populations. Geographic distance is a poor predictor of genetic distance between localities, especially towards the east of the range. This may be due to range alteration over the time-scale reflected by ISSR polymorphism. Inter-SSR variation declined from south to north in the western arm of the range, consistent with the prediction of Holocene aridification starting first and being most extreme in the north. Areas shown by the marker to harbour populations with high levels of variability include most parts of the eastern arm of the range, and the Kamiesberg highlands. The present study constitutes a demonstration of the range of evolutionary questions that can be addressed using a range of molecular phylogenetic and population genetic techniques, and the elucidation of both deep and shallow evolutionary history of a single Cape floral lineage.

Includes bibliographical references (p. 161-194).