TreeBASE Search-Matrices for study 311

@ARTICLE{TreeBASE2Ref14678, author = {Bernard R. Baum and D. B. O. Saville}, title = {Rusts (Uredinales) of Triticeae: Evolution and extent of coevolution, a cladistic analysis.}, year = {1985}, keywords = {}, doi = {}, url = {}, pmid = {}, journal = {Botanical Journal of the Linnean Society}, volume = {91}, number = {}, pages = {367--394}, abstract = {Established evolutionary trends in Uredinales as a whole are reviewed, and primitive and advanced characters are presented. The rusts of all Poaceae are presented in a table that strongly indicates Bambusoideae to be the oldest and Pooideae the youngest subfamily. The rusts of Triticeae and their ecogeography are outlined; the rusts of Cyperaceae, selected as an out-group, are briefly summarized; and the available characters and character states for rusts of Triticeae are given. Host alternation complicates the analysis. The aecial host (never a grass) is ecologically associated with the unrelated telial (grass) host. There are no appropriate methods to permit analysis of the combined components: aecial host evolution, telial host evolution, rust evolution, and their coevolution. Also, several aecial hosts are unknown. Consequently it was necessary to omit aecial hosts from the analysis. Cladistic analysis of the rusts of Triticeae was performed using five methods and consisted of cycles of tree analysis and modification of character state trees. A cladogram put together from a Dollo and a Wagner cladogram was used as a basis for the classification of rusts given. Subsequently a cladistic analysis of genera of Triticeae, using presence/absence of rusts as characters (Brooks' approach) was performed. The Triticeae cladogram of Baum (1983) was also analysed. Distances between the cladogram generated by various methods and that of Baum were computed for each possible pair, using the method of Robinson & Foulds, and then the resulting distance matrix was reduced in dimensionality by principal components and non-metric multidimensional scaling. The results are discussed in light of the limitation of the analyses and the data. It is concluded that coevolution is limited and that frequent jumps to ecologically associated hosts explain the parallelism in evolution of rusts on Triticeae.} }