THE IMPACT OF MOLECULAR SYSTEMATICS ON HYPOTHESES FOR THE EVOLUTION OF ROOT-NODULE SYMBIOSES AND IMPLICATIONS FOR EXPANDING SYMBIOSES TO NEW HOST-PLANT GENERA

Current taxonomic schemes place plants that can participate in root no dule symbioses among disparate groups of angiosperms. According to the classification scheme of Cronquist (1981) which is based primarily on the analysis of morphological characters, host plants of rhizobial sy mbionts are placed in subclasses Rosidae and Hamamelidae, and those of Frankia are distributed among subclasses Rosidae, Hamamelidae, Magnol iidae and Dilleniidae. This broad phylogenetic distribution of nodulat ed plants has engendered the notion that nitrogen fixing endosymbionts , particularly those of actinorhizal plants, can interact with a very broad range of unrelated host plant genotypes. New angiosperm phylogen ies based on DNA sequence comparisons reveal a markedly different rela tionship among nodulated plants and indicate that they form a more coh erent group than has previously been thought (Chase et al., 1993; Swen sen et al., 1994; Soltis et al., 1995). Molecular data support a singl e origin of the predisposition for root nodule symbiosis (Soltis et al ., 1995) and at the same time support the occurrence of multiple origi ns of symbiosis within this group (Doyle, 1994; Swensen, 1996; Swensen and Mullin, In Press).