Kniphofia, a genus of approximately 71 species, is almost entirely African with two species occurring in Madagascar and one in Yemen. Commonly known as ‘red hot pokers’ they are popular among horticulturists. The genus is also well known for its complex alpha taxonomy. To date, no studies have examined the phylogenetic relationships among species or the evolutionary history of the genus, and little work has been done on their biogeography. The main focus of this study was (i) to review the alpha taxonomy, (ii) to assess diversity and endemism in Kniphofia, (iii) to use DNA sequence data to reconstruct a specieslevel phylogeny to understand intra-generic species relationships and evolutionary processes (iv) to use phylogeographic approaches to study the biogeography and evaluate biogeographical patterns, and (v) to assess anatomical variation and determine if anatomical characters are useful for species delimitation. It was found that the genus has six centres of diversity, five of which are centres of endemism. The South African Centre is the most speciose and is also the largest centre of endemism. Kniphofia shows a strong Afromontane grassland affinity in Tropical and East Africa. In South Africa, it is found from high altitudes to coastal habitats, with the most speciose regions being Afromontane grasslands. It is thus not considered to be an Afromontane element, but rather an Afromontane associate. Five major evolutionary lineages were identified using cpDNA sequence data (trnT-L spacer), four of which are southern African. The fifth lineage is represented by material from Madagascar, East and Tropical Africa. The nuclear ITS region failed to provide resolution, as many sequences were identical. The five lineages recovered using cpDNA showed some congruence with geographic origin rather than the taxonomic arrangement based on morphology. All of the species with multiple samples were non-monophyletic. This could be due to hybridisation and/or incomplete lineage sorting. The nested clade analysis, although preliminary, did not completely agree with the phylogenetic analyses. One of the three third level nested clades appears to show fragmentation between the Cape Region, KwaZulu-Natal and northern parts of southern Africa. Furthermore, another nested clades recovered suggest a range expansion and radiation from the Drakensberg into the adjacent Drakensberg-Maputoland-Pondoland transition. Morphological species of Kniphofia exhibited substantial leaf anatomical variation and anatomical characters do not cluster samples into their morphological species. The anatomical results do not fit any geographic pattern, nor do they correspond to the lineages recovered using molecular markers or the nested clades. Leaf anatomical variation does not appear to be influenced by geographical or environmental factors. However, hybridisation may play a role but was not tested in this study. In light of the above findings it is proposed that the evolutionary and biogeographic history of Kniphofia is strongly linked to tectonic events, and Quaternary climatic cycles and vegetation changes. Tectonic events (viz. uplifts) may have resulted in vicariance events that may account for the five cpDNA lineages recovered in phylogenetic analyses, while Quaternary climatic cycles and vegetation changes may have had a more recent impact on evolution and biogeography. It is hypothesised that the ancestral area for Kniphofia was much more widespread when Afromontane grasslands were more extensive during cooler and drier glacial episodes. Kniphofia on the high mountains of Tropical and East Africa would have tracked Afromontane grasslands as they expanded their ranges in cooler periods. While during wetter and warmer interglacial periods Kniphofia would have retreated into refugia on the mountains of Tropical and East Africa, with no gene flow possible between these refugia. In South Africa, where latitude compensates for altitude, Kniphofia may have maintained a distribution that extended into the lowlands even during interglacials. A cyclic climate change hypothesis implies that populations of Kniphofia (at different phases of the climatic cycle) would have experienced periods of contractions and fragmentation followed by periods of range expansion and coalescence or secondary contact. Altitudinal shifting is proposed to be the most likely mechanism for fragmentation and range expansion, and would would possibly promoted hybridisation. Within the five lineages there is evidence for recent differentiation as the branch lengths are short, there are numerous nonmonophyletic species and numerous identical haplotypes (cpDNA and ITS) which collectively indicate a recent radiation in southern Africa. A recent radiation would also account for the taxonomic confusion and difficulty in differentiating morpho-species. These climatic events may also account for the substantial anatomical variation in southern African Kniphofia species.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:rhodes/vital:4220 |
| Date | January 2007 |
| Creators | Ramdhani, Syd |
| Publisher | Rhodes University, Faculty of Science, Botany |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis, Doctoral, PhD |
| Format | xxiii, 401 leaves, pdf |
| Rights | Ramdhani, Syd |
Page generated in 0.0021 seconds