| Efforts to resolve Darwin's "abominable mystery"--the origin of angiosperms-- have led to the conclusion that Gnetophytes are the sister group to Angiosperms. While many morphological characters appear to support this conclusion, it is not supported by molecular evidence from chloroplast and nuclear genes. This study adds a large data set from the slowly-evolving mitochondrial gene, coxI, including sequences from every family of gymnosperms. T to C transitions contribute most of the variation at first and second positions of coxI, and are often the result of a high level of RNA editing--editing allows "synonymous" base substitutions in the DNA that are corrected during mRNA processing. While conifer, angiosperm and cycad sequences appear to require substantial editing, Ginkgo and gnetophyte sequences require little if any; hence the data are analyzed with and without known and putative RNA editing sites. Although editing should not give misleading phylogenetic analyses, it allows T to C substitutions which might be the source of homoplasy in analyses of gymnosperm coxI sequences; also, some sequences may have been processed and reinserted into the genome, becoming processed paralogs. Sequences from all three plant genomic compartments (mitochondria: coxI; chloroplast: rbcL; nucleus: 18S rRNA) are shown to give congruent trees that conflict with the morphological "anthophyte" conclusion. In fact, coxI, 18S rRNA, and combined data (coxI, 18S, and rbcL) parsimony trees have a highly supported alternate topology: a monophyletic gymnosperm clade is sister to the angiosperms. Maximum likelihood and distance trees from all three genes indicate that gnetophytes are sister to the conifers, concurring with some morphological characters, and suggesting that certain characters supporting the anthophyte clade might represent convergences or symplesiomorphies. An initial angiosperm-gymnosperm split supports a pre-Cretaceous angiosperm origin and implies that there is no one extant seed plant group sister to the angiosperms. The relationships among the families of extant conifers have also been elusive, and data from coxI suggests that the oldest extant conifer groups are the Pinaceae, the Podocarpaceae and the Araucariaceae. |
