I am also wondering about using the octoploid and crossing it with tetraploids. There are some very interesting tetraploid species roses like R. laxa and R. spinosissima altaica. What is the probability of creating hexaploids by crossing R. acicularis acicularis with R. laxa or R. spinosissima and then using these to create tetraploids using any of the interesting diploids? Is this even likely? Doing the simple math would imply so but I have no idea if the possibility of these types of crosses is actually plausable. Has anyone worked with something like this? Has there been any interesting results? Just wondering what kind of species combinations could be concocted.