Theor Appl Genet (1996) 92:273-277
Using RAPDs to study phylogenetic relationships in Rosa
T. Millan, F. Osuna, S. Cobos, A. M. Torres, J. I. Cubero

Fig. 2 Phenogram derived from analysis of 20 accessions of 17 Rosa species using 10 random primers.

Finally, R. banksiae and R. cymosa were placed, independently, very distantly from the rest of the species (Fig. 2). The position of R. cymosa, which is considered to be a relative of banksiae (Bean 1970), did not fit the botanical classification of the genus. Whether this position is an artifact or, whether both species are more alike from a morphological point of view than from a molecular one cannot be clarified from our results.

CybeRose note: This sort of report puts the techniques in doubt. If the "random" primers were selected to give good resolution among European species, that's fine. But when the same primers separate such closely allied species as R. banksiae and R. cymosa, we might wish for a different set of primers.