JARS v60n3 - Research Review: Rhododendron Taxonomy

Research Review: Rhododendron Taxonomy

The understanding of evolutionary relationships among the species in the genus Rhododendron moved forward another step with the publication of the paper "The Molecular Systematics of Rhododendron (Ericaceae): A Phylogeny based upon RPB2 Gene Sequences by Loretta Goetsch, Andrew Eckert and Benjamin Hall. Although the research results were published in 2005 1 , and many ARS members are familiar with the work, a summary is presented here to enable members who are not familiar with the research to keep abreast of ongoing taxonomic work using DNA.
In their research the authors used the RPB2 gene in eighty-seven species of Rhododendron, recovering, sequencing and analyzing it to determine whether the sections and subgenera in the current Rhododendron system of classification, the Edinburgh Revision, are monophyletic groups. A monophyletic group is an assemblage of species a)that shares a common ancestor and b) that includes all living descendants of that ancestor. In addition, the authors attempt to resolve relationships among Rhododendron sections, including Ledum, Menziesia, and the monotypic subgenera Candidastrum and Mumeazalea.
The current taxonomy for Rhododendron, which is based on the work of Hermann Sleumer and David Chamberlain, is morphological. It was derived by comparing, species by species, the characteristics of homologous plant parts, such as the leaf, the inflorescence or the seed. In contrast, the taxonomic system proposed by authors Goetsch, Eckert and Hall is based on the characters present at homologous positions in the same gene, more particularly whether that position contains an A, G, C or T in the chemical structure of the gene.
Two advantages of this method, are: 1) the number of different characters that can be scored by DNA sequencing far exceeds the number of morphological characteristics of a plant, thus increasing resolution, and 2) the evolutionary history can be deduced from the pattern of relationships between different DNA sequences.
In their research, once the authors obtained the results of the DNA sequencing and lined up all the sequences with homologous positions juxtaposed, a computer algorithm (parsimony or maximum likelihood) analyzed the data to generate an evolutionary - or phylogenetic - tree. This shows the relationships of the various species to one another, as do the branches of a human family tree. A statistical procedure called bootstrapping is used to measure the degree of certainty supporting each of the branches in this tree. In the one inferred by Goetsch, Eckert and Hall, all but one of the major branches had very high statistical support (low uncertainty).
So what exactly did the authors find? Their results showed that all rhododendrons in the current classification, with the exception of R. camtschaticum , fall within three "clades" or monophyletic groups. They found the current subgenus Rhododendron to be monophyletic and represented by "Clade A." It includes current sections Rhododendron, Pogonanthum and Vireya. The former genus Ledum is included in Clade A.
"Clade B" includes the current subgenus Hymenanthes plus the current section Pentanthera and R. canadense . (In the Edinburgh Revision, section Pentanthera was included in the subgenus Pentanthera.) The remaining taxa of subgenus Pentanthera are moved to Clade C.
"Clade C" includes the current sections Azaleastrum, Tsutsusi, Brachycalyx, Rhodora, Sciadorhodion, and Nipponicum as well as the subgenera Mumeazalea and Candidastrum. The former genus Menziesia is included in Clade C.
The authors propose a revised taxonomic system that reduces the number of subgenera from eight to six. Subgenus names Rhododendron, Hymenanthes, and Azaleastrum are retained, although the latter is considerably broadened in content. It is proposed that the subgenus designations Pentanthera, Candidastrum and Mumeazalea be discontinued, since the species within them are reassigned. Additionally, section Choniastrum is considered a separate subgenus, since its species are the only ones in Clade A that have no lepidote scales.
Two other papers have been published with results that are largely consistent with this paper:
Kurashige, y et al. Plant Systematics and Evolution, 2001, 228:1-14; and Gao, L.M. et al. Acta Botanica Sinica, 2002, 44:1351-1356.

S. N.

1 Published in Systematic Botany, 30(3): 616-626, 2005.