We are using nucleotide sequence data from three chloroplast genes (atpB, matK and rbcL) as well as chloroplast genome structural mutations to study evolutionary relationships in the Campanulaceae sensu stricto. Our work to date focuses on 18 representative taxa. In addition to elucidating phylogenetic relationships among the organisms, we are interested in investigating genome structural evolution in the family. Approximately seventy-two different structural mutations have been identified in the eighteen genomes including examples of inversions, gene losses, duplications, insertions and possibly transpositions. We will use the three nucleotide sequence data sets to develop a hypothesis of Campanulaceae phylogeny and then use the DNA sequence results to develop models of chloroplast genome structural evolution. Relative to most other land plants, Campanulaceae chloroplast genomes have experienced an unusually high amount of structural change. This presents an opportunity to better understand the mechanisms leading to structural mutation. In addition, Campanulaceae provides an opportunity to compare the effectiveness of structural data versus sequence data in phylogeny reconstruction. In our initial comparisons, consistency indices were higher in the structural data regardless of the method of character scoring or weighting used in parsimony analyses. Other measures also favored structural data. Plus the structural data produced similarly resolved trees. In Campanulaceae, the structural data provide many characters without high levels of homoplasy. Our comparisons suggest that chloroplast genome structural data is an important adjunct to other data sets in groups where such markers are available.

Key words: Campanulaceae, chloroplast genome, phylogenetic methods, structural mutation