Inferring plant evolutionary history from molecular data

Abstract

Abstract The rapid development of molecular methods during the 1980s has had a profound effect on the study of plant evolution. Molecular data have accumulated very rapidly, and this abundance of data poses new problems for data analysis. We illustrate three problem areas that arise in plant evolutionary inference. The first problem concerns the use of molecular data to analyse closely related plant species linked through reticulate evolution. The second problem arises from our ability to sample DNA sequences from alleles of a single genetic locus within plant species. We show how cumulative information on selection and random genetic drift can be extracted from such data. The third problem area concerns the limitations of current algorithms for phylogenetic inference when confronted with large sets of DNA sequence data. Based on a consideration of these problem areas, we conclude that: (1) asymmetric transmission of cpDNA markers is useful in resolving the parentage of hybrid plant taxa; (2) simple clustering algorithms can provide useful information on cultivar or variety relationships, despite intervarietal hybridisation, if genetic similarities are averaged over sufficient loci; (3) samples of complete DNA sequence data from plant nuclear genes can provide a new dimension of information on historical effective population sizes and on the mechanisms that generate allelic diversity; (4) analyses of the chloroplast gene rbcL, sampled from across the monocotyledon class, reveal large variation in relative rates of nucleotide substitution—these variations, in turn, have important consequences for phylogeny estimation algorithms; and (5) the combined use of algorithms like parsimony and maximum likelihood may represent a more efficient approach to the phylogenetic analysis of very large (>100) sets of DNA sequences.