A phylogeny of Begoniaceae Bercht. & J. Presl

Forrest, Laura Lowe (2000) A phylogeny of Begoniaceae Bercht. & J. Presl. PhD thesis, University of Glasgow.

Full text available as:
[img]
Preview
PDF
Download (17MB) | Preview

Abstract

Begonia is one of the largest angiosperm genera, with 1400 species currently recognised. These were placed into 63 sections in the most recent taxonomic treatment. However, there is considerable uncertainty in both section inter-relationships and sectional composition, and there is no formalised phylogenetic hypothesis for the genus. Using the nuclear internal transcribed spacers (ITS) and partial large subunit (26S) sequences of ribosomal DNA, I have produced phylogenetic trees to form the basis of a cladistic framework for the interpretation of the evolution and sectional level systematics of Begonia. Maximum parsimony, maximum likelihood and minimum evolution cladograms were produced for 35 Begonia, one Symbegonia and two Datisca species, for partial 26S, ITS and combined sequence data. The results of the analyses suggest that African taxa are basal in Begonia, but that there is not sufficient information to elucidate the precise relationships among these basal lineages. The genus Symbegonia is nested deeply within Begonia. A far larger data set was constructed by sequencing the ITS region for 153 species. Different alignment methods (automated, elision and manual) were tested on these sequences, as were different search strategies. The topology which was taken to be the best estimate of the ITS phylogeny of Begoniaceae was constructed using manual alignment, culled of ambiguous regions, and adjusted to reflect the topologies of smaller, localised, compartment analyses. In the resulting tree, the African species of Begonia resolve as paraphyletic, with both Asian species (including Socotra) and American species (sister to southern African species) monophyletic. Comparisons were made between the ITS sequence data and trees produced from the chloroplast trnC - trnD inter-genic spacer. Parsimony analyses of trnC - trnD sequences support African taxa as basal in Begonia] however, in contrast to the ITS data, trnC - trnD suggests polyphyly / paraphyly of American taxa, albeit with little bootstrap support. A morphological data matrix (67 characters for 159 taxa) did not produce a phylogenetic hypothesis for Begonia that was congruent with any other available data. Using a combined morphology - ITS analysis, the fit of individual morphological characters to a fundamental tree was examined. Some characters fitted the combined topology well, although some of the characters which have traditionally been considered important in Begonia taxonomy (e.g. the number of placental branches) proved misleading. The ITS tree was used as a framework for reviewing chromosome evolution in the genus (604 published counts from 239 species). In contrast to sequence divergence, which was greatest among African species, chromosome number diversity was greatest among American species. The correlation between phylogenetic relationships implied by the ITS tree and the geographical distributions of species was explored to obtain biogeographic hypotheses which may explain the present-day distribution of Begonia. As a general rule, related species are geographically proximal, suggesting limited dispersal of lineages. This finding contradicts observations made on morphology, where the close affinities of morphologically disparate (but geographically proximal) taxa were previously unsuspected. Mechanisms responsible for the evolution of large genera were discussed, and Willis' 'age and area' hypothesis compared to the 'relict' hypothesis of Cronk. In Begonia, the morphological diversity of the genus, and most of the species, are encompassed among the putatively derived lineages, favouring the 'relict' hypothesis.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Additional Information: Adviser: Pete Hollingworth
Keywords: Botany
Date of Award: 2000
Depositing User: Enlighten Team
Unique ID: glathesis:2000-73472
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 14 Jun 2019 08:56
Last Modified: 14 Jun 2019 08:56
URI: http://theses.gla.ac.uk/id/eprint/73472

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year