Guioa cav. (Sapindaceae): Taxonomy, phylogeny, and historical biogeography

A monograph of the genus Guioa Cav. (Sapindaceae) is presented. 64 species are recognized; one species, Guioa subfalcata, is regarded as a dubious species. A general key and several regional keys provide access to the species. One species, Guioa glauca, is subdivided into two varieties. The infrageneric classification contains three levels: subgenera, sections, and subsections. The classification is based on a phylogenetic analysis of the genus. The leaf anatomy was studied besides the macromorphology. This resulted in 67 characters for a cladistic analysis. The outcome of the initial analysis was rejected for several reasons (chapter 11.6). An alteration of the method resulted in an acceptable cladogram: the genus was subdivided, based on several characters, into 5 groups, 4 of which were subordinate to each other. Each subgroup was analysed separately and contained a representative species for a higher group and an outgroup which was obtained from a lower group. The cladograms of the 5 subgroups were pieced together to form the accepted cladogram. The computer programs HENNIG86 and PAUP provided the best method for a cladistic analysis (Wagner algorithm combined with some kind of ‘branch and bound’-option). A historical biogeographic analysis was performed with two aims: a) to find the historical relationships between the different distribution areas and b) to test the generic cladogram by comparing it with cladograms of independent groups of plants and animals. Like with the genus, the areas were subdivided into several groups, which were analysed separately under assumption 0 via a parsimony method instead of a consensus method. The complete generalized areagram was later on pieced together. The partial generalized areagrams did not falsify the accepted cladogram of Guioa. In two areas, the Pacific and West Malesia, Guioa mainly showed a pattern of dispersal and not of vicariance. The computer program CAFCA (group- and component-compatibility) presented the best method for a historical biogeograpic analysis. A comparison between the presumed results of the classical, evolutionary school and the phylogenetic school showed that the results of the phylogenetic school had a higher information content in the sense of natural relations and that only they could be used for a historical biogeographic analysis. The ecological part contains data on the habitat of some Guioa species, a compilation of mainly literature data on floral biology of Sapindaceae in general and on fruit biology, data on seed germination and seedling morphology. In order to unravel the species complex ( Guioa rigidiuscula group) on New Guinea only an artificial technique could be used to separate the 14 species. These appear to be natural due to correlating characters and coherent, non-disjunct distributions. It is discussed that species complexes are mainly found in especially geologically dynamic areas and that the mode of speciation is then best explained by the model of punctuated equilibria.