Floriculture and Ornamental Biotechnology (2009)
The Inheritance of Juvenile Recurrence in Rosa Species Hybrids
Roger E. Mitchell II*

ABSTRACT
The current market for roses demands that new cultivars bloom throughout the growing season. The form of recurrence found in most modern roses can be termed juvenile recurrence, because bloom begins only months after germination. Previous studies have observed that juvenile recurrence is conferred by a recessive allele at a single locus. Most species in the genus Rosa have not yet been used to produce commercialized hybrids, in part because they carry the dominant allele for non-recurrence. Furthermore, only a few species have been tested for their effect on recurrence when used for hybridization. This study investigated the inheritance of juvenile recurrence by crossing several hardy tetraploid species and near-species hybrids with modern roses, and then backcrossing the resulting hybrids with modern roses to recover recurrence. For all species, and all but one first-generation hybrid, juvenile recurrence was recovered. This suggests that using non-recurrent species in a rose breeding program is feasible. The numbers of recurrent and non-recurrent second-generation backcross offspring produced by each first-generation hybrid varied. Some progeny groups did not differ significantly from the theoretically predicted 1 recurrent: 5 non-recurrent ratio, while the number of recurrent offspring was lower than predicted in the rest. The ability of some species and near-species hybrids to produce some late blooms did not affect the frequency of juvenile recurrence in the second generation.


... in diploid crosses of R. wichurana [sic] and a hybrid China rose ('Old Blush'), Shupert and Byrne (2007) saw greatly reduced ratios of recurrent to non-recurrent offspring. Svejda (1976) also saw low percentages of recurrence in diploid crosses of recurrent hybrid China rose cultivars with selected forms of the recurrent species R. rugosa Thunb. ex. Murray. In this case, only 28% of the offspring were recurrent, much lower than the 100% predicted by the single-locus theory. Taken collectively, these studies indicate that the single-locus, r/R theory, while generally accurate, can be complicated by other factors.


Variable forms and expression of recurrence

The issue of recurrence is further complicated by the apparent existence of more than one form of expression. The work of Svejda (1976) suggests that forms and hybrids of R. rugosa gain recurrence from the same r allele as modern roses. Despite this, the expression of recurrence is different. Rosa rugosa does not usually begin to bloom until plants have grown to nearly mature size (Zlesak 2001). Late-season bloom has also been described in a number of other wild species roses, including R. laxa Retzius (Buck 1962) and R. arkansana (Erlanson 1938; Collicutt 1992). Much work remains to be done to determine which, if any, of these species are genuinely recurrent, whether they share the same mechanism of recurrence as modern roses, and whether their recurrence can be utilized to simplify the task of breeding species hybrids possessing juvenile recurrence.