Phylogenetic analysis of green plant rbcL sequences

Abstract

DNA and derived amino acid sequences of the rbcL gene from selected photosynthetic bacteria, green algae, bryophytes, "fern allies," ferns, and seed plants were used to construct phylogenetic trees using parsimony algorithms. Internal support of clades was determined by the bootstrap and decay indices. External support was judged by comparison of trees with previous morphological and molecular cladograms. Some clades on all the trees are supported both internally and externally, but all of the trees are characterized by unsupported relationships of the bryophytes and fern allies. Trees produced by analyses of all three DNA positions have the highest levels of internal support, but many clades are still not supported externally or internally. Analyses using only the first and second positions of the DNA sequences and derived amino acid sequences produced trees with odd groupings and low levels of internal support for most of the clades. These results indicate that rbcL sequences may be too divergent to test phylogenetic relationships among major groups of green plants. Additional problems with constructing phylogenetic trees from rbcL sequences may be caused by RNA editing, pseudogenes, unequal rates of evolution, and inadequate taxon sampling. Whether RNA editing is a problem can be determined by sequencing amplified RNAs from selected taxa. If pseudogenes are suspected, purified chloroplast DNA should be used for amplification. It may be beneficial to include more taxa in subsequent analyses, but there are indications that historical extinctions will prevent adequate sampling of some groups, such as the bryophytes and fern allies.