CHLOROPLAST DNA VARIATION WITHIN AND AMONG GENERA OF THE HEUCHERA GROUP (SAXIFRAGACEAE): EVIDENCE FOR CHLOROPLAST TRANSFER AND PARAPHYLY

The Heuchera group (Saxifragaceae) comprises Bensoniella, Conimitella, Elmera, Heuchera, Lithophragma, Mitella, Tellima, Tiarella, and Totmiea. Earlier studies employing morphology, karyology, and flavonoid chemistry indicated that these genera form a natural group, but failed to resolve relationships among them. Restriction site analysis of chloroplast DNA (cpDNA) suggests that Bensoniella, Tolmiea, and Lithophragma are close allies and form the sister group of a large clade containing the remaining six genera. Mitella and Heuchera are both paraphyletic based on cpDNA data. cpDNA data, in conjunction with morphological and allozyme data, suggest at least four examples of intersectional hybridization and subsequent chloroplast capture in Heuchera. Several of these events may be explained via a stepping stone model in which the chloroplast genome of a species was captured by a second species, and then ultimately by a third taxon. Two well-differentiated groups of Tellima populations were detected: one group has a unique chloroplast genome characterized by nine autapomorphies, and the second group has a chloroplast genome identical to that found in M. trifida and M. diversifolia. cpDNA and allozyme data suggest that some Tellima populations probably obtained their chloroplast genome via intergeneric hybridization with M. trifida, M. diversifolia, or the ancestor of these taxa. The occurrence of intergeneric chloroplast transfer in some populations of Tellima, as well as extensive intersectional chloroplast capture in Heuchera, not only suggests caution in the use of cpDNA restriction site data in phylogenetic reconstruction, but also demonstrates again the importance of adequate sampling of conspecific populations. If the intergeneric relationships in the Heuchera group suggested by cpDNA analysis are accurate, fundamental questions arise regarding the validity of certain morphological traits as good taxonomic characters in Saxifragaceae. Furthermore, significant taxonomic changes at the generic level would be necessary.