Rosales sister to Fabales: Towards resolving the rosid puzzle

Abstract

Rosids represent the largest of the eight major clades of core eudicots comprising of 140 families and approximately one-third of all angiosperm species (Soltis and Soltis, 2004). Other than the traditional Rosidae, it encompasses families from Magnoliidae, Dilleniidae and Hamamelidae (Cronquist, 1981; Takhtajan, 1980, 1997). Relationships within the rosids still remain unclear, but with addition of new data in the form of complete plastome sequences, pieces of the puzzle seem to be falling in the right places. With the completion of the grape chloroplast genome (Jansen et al., 2006), doubts regarding “Vitis sister to the rosids” have been eliminated. However, relationships within the two large subclades of rosids, eurosids I (fabids) and II (malvids), remain uncertain and poorly resolved (Savolainen et al., 2000a,b; Soltis et al., 2000). Eurosids I comprises of orders Celastrales, Oxalidales, Malpighiales, Zygophyllales and the nitrogen-Wxing clade comprising of orders Cucurbitales, Fabales, Fagales and Rosales (Soltis et al., 1995). Eurosids II comprises of orders Brassicales, Malvales and Sapindales. The rosids remain one of the most challenging problems for molecular phylogeneticists within the angiosperms (Soltis and Soltis, 2004). Single/few gene analyses have not been able to provide robust support to these problematic areas. In a study on Urticalean rosids, Sytsma et al. (2002), using rbcL, trnL-F and ndhF sequences, revealed that Fagales was sister to Rosales. Fabales appeared sister to Fagales + Rosales and Cucurbitales sister to (Fabales, (Fagales, Rosales)). However, the branch points received low bootstrap support