Angiosperm phylogeny inferred from 18S ribosomal DNA sequences

Abstract

Parsimony analyses were conducted for 223 species representing all major groups of angiosperms using entire 18S ribosomal DNA (rDNA) sequences. Although no search swapped to completion, the topologies recovered are highly concordant with those retrieved via broad analyses based on the chloroplast gene rbcL. The general congruence of 18S rDNA and rbcL topologies further clarifies the broad picture of angiosperm phylogeny. In all analyses, the first-branching angiosperms are Amborellaceae, Austrobaileyaceae, Illiciaceae, and Schisandraceae, all woody magnoliids. These taxa are always followed by the paleoherb family Nymphaeaceae. This same general order of early-branching taxa is preserved with several suites of outgroups. In most searches, the remaining early-branching taxa represent Piperales and other orders of subclass Magnoliidae (sensu Cronquist). With the exception of Acorus, the monocots are supported as monophyletic and typically have as their sister Ceratophyllum. In most analyses, taxa with uniaperturate pollen form a grade at the base of the angiosperms; a large eudicot clade is composed primarily of taxa having triaperturate pollen. Two large subclades are present within the eudicots, one consisting largely of Rosidae and a second corresponding closely to Asteridae sensu lato. Subclasses Dilleniidae and Hamamelidae are highly polyphyletic. These data sets of 18S rDNA sequences also permit an analysis of the patterns of molecular evolution of this gene. Problems deriving from both the prevalence of indels and uncertain alignment of 18S rDNA sequences have been overstated in previous studies. With the exception of a few well-defined regions, insertions and deletions are relatively uncommon in 18S rDNA; sequences are therefore easily aligned by eye across the angiosperms. Indeed, several indels in highly conserved regions appear to be phylogenetically informative. Initial analyses suggest that both stem and loop bases are important sources of phylogenetic information, although stem positions are prone to compensatory substitutions. Of the stem changes analyzed, only 27% destroy a base-pairing couplet; 73% maintain or restore base pairing.