They didn’t stratify the seeds. After they softened the pericarp with hydrochloric acid, they removed the pericarp and testa from some achenes, and left it on others. The achenes whose pericarps and testas had been removed germinated immediately. The other achenes did not germinate.
For some kinds of seeds, imbibition of water is sufficient to break dormancy, but rose achenes are impermeable to water. The researchers apparently wanted to find out whether they could break dormancy by using acid to make the achene permeable to water. Water permeability was insufficient, apparently because there was still ABA in the pericarp.
Their suggestion that leaching the ABA out of the softened achenes would stimulate germination seems reasonable. It is frustrating that they didn’t do the experiment that would prove it.
At least they demonstrated a way to remove the pericarp from achenes without damaging the embryos. Last year, I tried several ways to remove pericarps, but damaged so many of the embryos that I had to give up on it.
We should note that the experiment was performed on seeds of only one cultivar, and cannot be taken as a blanket recommendation for all roses. We know, for example, that in some crosses the pollen parent has a major influence on seed dormancy, which implies that the dormancy resides (in part) in the embryo/endosperm of such seeds – rather than just in the testa and pericarp as in ‘Crimson Glory’.
Then, too, some seeds require prolonged heat AND prolonged cold – which implies two forms of dormancy that have to be broken independently. Rosa canina seeds germinate very readily (so I’ve read) if they are kept warm (greenhouse temperature) for a month before receiving a month of cold. The seedlings came up like radishes. Other seeds that were merely chilled germinated poorly – some waiting for a second or third year.
I’d really like to learn more about the Pimpinellifoliae species. Some (e.g. omeiensis) ripen their seeds and hips in early summer, and presumably have a dormancy that is broken by heat in addition to the familiar winter chilling requirement. This may explain Peter Harris’s observation that ‘Golden Showers’ seeds germinate mostly in the autumn after they ripen – winter chilling followed by a summer roasting.
Ivan Michurin, the great Russian plant breeder, wrote that he always cracked rose seeds (achenes) before planting them. That would avoid the impermeable pericarp problem.
Lucian Reychler (Belgian) peeled the testa from some very old peach seeds and got rapid germination – much faster than from fresh seeds. This might not be too bad if we have only a handful of seeds.
There are no doubt variations on the dormancy mechanism in different rose species. ABA probably plays a central role in all of them.
There are at least two parts to the ABA mechanism in ‘Crimson Glory’, the quantity of ABA in the pericarp and testa, which is controlled by the seed parent, and the sensitivity of the embryo to ABA, which is controlled by both parents. The pollen parent would influence the embryo’s sensitivity to ABA. The pollen parent could also influence the amount of ABA in the embryo.
The ‘Crimson Glory’ embryos in the study contained very low amounts of ABA. I would expect that the embryos of some rose species would contain higher amounts of ABA. Those species could require as many as three different things to happen to break dormancy: a decrease in ABA levels in the pericarp and testa, a decrease in ABA levels in the embryo, and a decrease in ABA sensitivity in the embryo. Different environmental factors could control each of these three things.
Of course, this is all just speculation since rose achene dormancy is not well understood, but it is consistent with what is known about dormancy in other kinds of seeds. For example, yellow cedar seeds need a period of warm stratification followed by a period of cold stratification to germinate (see link). Yellow cedar embryos have high levels of ABA and the embryos have a high sensitivity to ABA. Most of the ABA in the embryo is lost during the warm stratification, but the subsequent cold stratification is required to reduce the embryo’s sensitivity to ABA. The authors conclude, “A decline in ABA within the seed is not sufficient for dormancy breakage; reduced embryo sensitivity to ABA is also required.” It would be interesting to know whether ABA levels are higher in R. canina embryos than ‘Crimson Glory’ embryos. It might help to explain why R. canina requires warm stratification while ‘Crimson Glory’ doesn’t.