Phylogenetic relationships amongst the African genera of subtribe Orchidinae s.l. (Orchidaceae; Orchideae): Implications for subtribal and generic delimitations.

Phylogenetic relationships within the Orchideae sensu Pridgeon et al, remain one of the biggest unresolved issues in our understanding of the taxonomy of the orchids. Members of the Orchideae are numerous and widespread in Africa but remain poorly represented in phylogenetic research. In this study we included a broad sampling of African taxa for which we sequenced three plastid (rbcl, matK and trnL + trnL-F) and two nuclear regions (ITS and 18S). We used 368 sequences representing 278 species and 49 genera to infer relationships using the Bayesian Inference and Maximum Likelihood method. Our results show strong support for three clades, two of which almost entirely match the historical circumscription of Orchidinae and Habenariinae, and the third, Bartholininae, sister to the former two, includes the genera Holothrix and Bartholina. Stenoglottis should be assigned to Orchidinae and not to Habenariinae. Several genera such as Habenaria, Cynorkis and Benthamia are shown to be para- or polyphyletic: Bonatea, Centrostigma, Platycoryne and Roeperocharis are all embedded in Habenaria; Physoceras, Arnottia and part of Benthamia are embedded in Cynorkis. We propose a subdivision of Orchideae sensu lato into nine subtribes, but refrain from making generic re-arrangements until more extensive or more in-depth studies have been done.

An approach to identify putative hybrids in the 'coalescent stochasticity zone', as exemplified in the African plant genus Streptocarpus (Gesneriaceae).

The inference of phylogenetic relationships is often complicated by differing evolutionary histories of independently-inherited markers. The causes of the resulting gene tree incongruence can be challenging to identify, often relying on coalescent simulations dependent on unverifiable assumptions. We investigated alternative techniques using the South African rosulate species of Streptocarpus as a study group. Two independent gene trees - from the nuclear ITS region and from three concatenated plastid regions (trnL-F, rpl20-rps12 and trnC-D) - displayed widespread, strongly supported incongruence. We investigated the causes by detecting genetic exchange across morphological borders using morphological optimizations and genetic exchange across species boundaries using the genealogical sorting index. Incongruence between gene trees was associated with ancestral shifts in growth form (in four species) but not in pollination syndrome, suggesting introgression limited by reproductive barriers. Genealogical sorting index calculations showed polyphyly of two additional species, while individuals of all others were significantly associated. In one case the association was stronger according to the internal transcribed spacer data than according to the plastid data, which, given the smaller effective population size of the plastid, may also indicate introgression. These approaches offer alternative ways to identify potential hybridization events where incomplete lineage sorting cannot be rejected using simulations.

Phylogeography and molecular evolution of potato virus Y.

Potato virus Y (PVY) is an important plant pathogen, whose host range includes economically important crops such as potato, tobacco, tomato, and pepper. PVY presents three main strains (PVY(O), PVY(N) and PVY(C)) and several recombinant forms. PVY has a worldwide distribution, yet the mechanisms that promote and maintain its population structure and genetic diversity are still unclear. In this study, we used a pool of 77 complete PVY genomes from isolates collected worldwide. After removing the effect of recombination in our data set, we used bayesian techniques to study the influence of geography and host species in both PVY population structure and dynamics. We have also performed selection and covariation analyses to identify evolutionarily relevant amino acid residues. Our results show that both geographic and host-driven adaptations explain PVY diversification. Furthermore, purifying selection is the main force driving PVY evolution, although some indications of positive selection accounted for the diversification of the different strains. Interestingly, the analysis of P3N-PIPO, a recently described gene in potyviruses, seems to show a variable length among the isolates analyzed, and this variability is explained, in part, by host-driven adaptation.

Genetic variability within the coat protein gene of Grapevine fanleaf virus isolates from South Africa and the evaluation of RT-PCR, DAS-ELISA and ImmunoStrips as virus diagnostic assays.

Grapevine fanleaf virus (GFLV) is responsible for severe fanleaf degeneration in grapevines of all major wine producing regions of the world, including South Africa. In order to successfully control the spread of the virus, specific and reliable diagnostic assays are necessary. The genetic variability of 12 GFLV isolates recovered from naturally infected grapevine plants in the Western Cape region of South Africa were characterised. These samples were subjected to RNA extraction, RT-PCR analysis and sequencing of the coat protein gene (2CCP). Sequence identities between different GFLV isolates from South Africa were between 86-99% and 94-99% at the nucleotide and amino acid levels, respectively. Phylogenetic analysis based on the 2CCP gene sequences showed that the South African isolates form two distinct clades or sub-populations. The specificity and sensitivity of three diagnostic techniques (rapid-direct-one-tube-RT-PCR, DAS-ELISA and ImmunoStrips) for the detection of GFLV were analysed to determine the appropriate diagnostic assay for virus infection. Rapid-direct-one-tube-RT-PCR was found to be the most reliable technique for detection. This is the first report on sequence analysis of full-length 2CCP gene cDNA clones of GFLV isolates from South Africa.

Origin and diversification of the Greater Cape flora: ancient species repository, hot-bed of recent radiation, or both?

Like island-endemic taxa, whose origins are expected to postdate the appearance of the islands on which they occur, biome-endemic taxa should be younger than the biomes to which they are endemic. Accordingly, the ages of biome-endemic lineages may offer insights into biome history. In this study, we used the ages of multiple lineages to explore the origin and diversification of two southern African biomes whose remarkable floristic richness and endemism has identified them as global biodiversity hotspots (succulent karoo and fynbos). We used parsimony optimization to identify succulent karoo- and fynbos-endemic lineages across 17 groups of plants, for which dated phylogenies had been inferred using a relaxed Bayesian (BEAST) approach. All succulent karoo-endemic lineages were less than 17.5 My old, the majority being younger than 10 My. This is largely consistent with suggestions that this biome is the product of recent radiation, probably triggered by climatic deterioration since the late Miocene. In contrast, fynbos-endemic lineages showed a broader age distribution, with some lineages originating in the Oligocene, but most being more recent. Also, in groups having both succulent karoo- and fynbos-endemic lineages, there was a tendency for the latter to be older. These patterns reflect the greater antiquity of fynbos, but also indicate considerable recent speciation, probably through a combination of climatically-induced refugium fragmentation and adaptive radiation.

A molecular phylogeny for the large African orchid genus Disa.

Phylogenetic relationships were inferred for the African subtribe Disinae (Orchidoideae, Orchidaceae), which include the large genus Disa and the small genus Schizodium. One nuclear (ITS) gene region and two plastid (trnLF and matK) gene regions were sequenced for 136 ingroup, representing 70% of all known Disinae species, as well as for 7 outgroup taxa. The combined data matrix contained 4094 characters and was analysed using parsimony and Bayesian inference. Our results show that the generic status of Schizodium can no longer be supported, as it is deeply embedded within the genus Disa. Furthermore, the currently recognised subgenera do not reflect the phylogenetic relationships and should be rejected. Several of the currently recognised sections are monophyletic, others contain misplaced elements, while some are polyphyletic. Morphological divergence, rather than convergence, has hampered previous attempts at a phylogenetic classification of the Disinae. On the basis of our molecular phylogenetic hypothesis, we propose a monotypic subtribe Disinae and a subdivision of the genus Disa into 18 sections.