Phylogeny and biogeography of Chara (Charophyta) inferred from 18S rDNA sequences

Competing hypotheses concerning the relationships among subsections of the genus Chara were tested with parsimony and maximum likelihood analysis of s mall subunit (18S) ribosomal DNA gene (rDNA) sequences. The traditional hyp othesis as slated by Wood and Imahori (1965) separates Chara into subgenera and subsections based primarily on stipulode and cortication characters, w hereas Proctor's (1980) hypothesis proposed subgenera based on the break-up of Pangea into Laurasia and Gondwana, and subsections based on the subsequ ent break-up of these two supercontinents. Both parsimony and maximum likel ihood analyses strongly refuted Wood and Imahori's hypothesis, as well as t he importance traditionally placed on stipulode and cortication characters. Partial or total loss of cortication may occur within a subsection, as dem onstrated by C. imperfecta, C. braunii, C. muelleri and C. brittonii. Altho ugh the nodal stem cell gives rise to the branchlets, stipulodes and axial cortex, stipulodes can be retained despite the loss of cortication as demon strated by C. hydropitys, which is placed solidly within Willdenowia by bot h analyses. Proctor's hypothesis was supported with Gondwanan subsections b asal (except for C. tomentosa) and Laurasian subsections placed together in a monophyletic clade in both analyses, although the placement of subsectio n Grovesia is uncertain. Monophyly of subsections Wallmania and Willdenowia and Lamprothamnium are strongly supported. Lamprothamnium is placed within the genus Chara, which disagrees with previous ((McCourt et al. 1996) rbcL analyses. Further studies are needed to identify phylogenetically informat ive morphological characters for extant taxa, and a method is needed to lin k extant species to the extensive fossil data for a fuller understanding of the phylogeny and biogeography of this distinctive genus.