Dr. Susan E. Meyer
Germination and seed testing.
Rose seeds are normally dormant at maturity and require some form of pretreatment in order to germinate. Release from dormancy is a complex process that may involve changes at the pericarp, testa, and embryo levels. The degree of dormancy and the principal level of dormancy control varies among species, cultivars, seedlots, and even among hips within a single bush. Because the achenes have a thick, hard pericarp and do not swell when placed in water, it is often assumed that they are water-impermeable. Work by Svejda (1972) and others has shown that this is not the case. The achenes do take up water, although the mechanical restriction presented by the pericarp can sometimes prevent full imbibition. Tincker and Wisley (1935) showed, for 10 rose species, that cracking the pericarp alone did not remove dormancy. The importance of including treatments that weaken the pericarp in efforts to remove rose seed dormancy depends on the species and the particular lot. In nursery propagation of the rootstock rose R. dumetorum (R. corymbifera) ‘Laxa’, sulfuric acid treatment before warm plus cold stratification improves germination (Roberts and Shardlow 1979). The acid scarification can be eliminated and the warm stratification period shortened if the achenes are warm-stratified with compost activator (Cullum and others 1990). The role of these treatments is apparently to weaken the pericarp along the sutures, whether with acid or through microbial digestion. Responsiveness to warm plus cold stratification can also be increased in R. dumetorum 'Laxa' by vacuum-infiltrating the achenes with growth hormones such as gibberellic acid or benzyladenine (Foster and Wright 1983), which suggests that something other than simple mechanical restriction may be involved.
Similarly, in the relatively non-dormant multiflora rose, the achenes may be induced to germinate without chilling either by treatment with macerating enzymes that weaken pericarp sutures or by leaching with activated charcoal to remove inhibitors from the incubation solution (Yambe and Takeno 1992; Yambe and others 1992). By using macerating enzyems to remove dormancy, these workers were able to demonstrate a phytochrome-mediated light requirement for germination in this species (Yambe and others 1995). Acid scarification (but not mechanical scarification) is reported to substitute for warm pretreatment in the cultivated rose R. gallica L (Svejda 1969).
Chilling is the treatment most often applied to remove rose seed dormancy, and the achenes of most species will germinate eventually if chilled for long enough periods. For some species, periods of cold stratification corresponding to a single winter in the field are sufficient, as in prairie, multiflora, and wichura roses (table 3). Achenes of these species may show increased dormancy if the chilling period is preceded or interupted by periods of incubation at warmer temperatures (Semeniuk and Stewart 1962; Stewart and Semeniuk 1965). Interruption of chilling with warm incubation resulted in secondary dormancy induction only if the temperature of warm incubation was too high. If the seeds were held below this 'compensating' temperture, no change in dormancy resulted, and the seeds could accumulate the effects of chilling across warm interruptions. Seeds whose chilling requirements had just barely been met germinated best at relatively low incubation temperatures, whereas those that had been in chilling for longer than necessary either eventually germinated in chilling or could germinate at a wide range of temperatures, including those above the compensating temperature. Semeniuk and others (1963) showed that, for prairie rose, the effect of the warm pretreatment  above the compensating temperature was to induce secondary dormancy at the embryo level. Interestingly, this dormancy could be alleviated only by chilling whole achenes: chilling the embryos did not alleviate their dormancy.
Other species, such as prickly, Nootka, and Woods roses, show much increased germination percentages in response to chilling periods corresponding to a single winter if the chilling period is preceded by a period of warm incubation (table 3). This requirement for warm incubation before chilling would effectively postpone seedling emergence in the field until the second spring after seed production (Densmore and Zasada 1977). The temperature and duration of the warm treatment is sometimes important. In rugosa rose, a warm pretreatment of 60 days at 20 °C before 90 days of chilling at 3 °C increased germination over chilling alone, but longer periods resulted in decreased germination (Svejda 1968). The effect of warm pretreatment on chilling response has been formally documented for only a few rose species. but it is likely that high-viability lots of any species that show minimal germination after 6 months of chilling would he benefitted by a warm pretreatment.
Exactly what changes take place in rose seeds during warm pretreatment or chilling is not known. In many cases, the warm pretreatment seems to have effects at the seed level rather than simply providing an opportunity for pericarp weakening (Densmore and Zasada 1977). Hormonal balance has been implicated in the imposition of dormancy in ruse seeds by several workers. Substances leached from dormant rose achenes or obtained from them by grinding have been shown to suppress germination of otherwise nondormant excised rose embryos (Jackson and Blundell 1963. 1965; Svejda and Poapst 1972). Excised seeds with physically disrupted testas showed much lower germination than embryos with testas removed, suggesting that inhibitors leaching from the testa suppressed germination (Jackson and Blandell 1963). Other workers have shown that, although inhibitory substances are present in dormant achenes and may disappear during dormancy loss, their removal alone is not sufficient to induce germination (Julin-Tegelman 1983; Tillberg 1983).
Variation in dormancy-breaking requirements both within and among lots of any rose species make it difficult to predict effective treatments. One of the causes of this variation has been quite well-studied in cultivated tea roses. and the results probably apply to wild species as well. Von Abrams and Hand (1956) were the first to demonstrate that seeds of a given cultivar matured in the field at warmer temperatures were less dormant (that is. had a shorter chilling requirement) than seeds matured at cooler temperatures. This result has been confirmed by De Vries and Dubois (1987), who also found that warmer maturation temperatures were associated with higher hip set and higher numbers of achenes per hip. Gudin and others (1990) examined the relationship of maturation temperature with developmental rate, endocarp thickness, and dormancy status.. They also looked at the effect of the pollen parent in controlled crosses. They found that achenes matured at cooler spring temperatures had slower development, thicker endocarps. and higher levels of dormancy than those matured at warmer summer temperatures. Pollen parent also had an effect on both dormancy and endocarp thickness, presumably through its effect on developmental rate. These workers concluded that the higher dormancy associated with lower maturation temperature was mediated through endocarp thickness, but slow development could also have effects at the testa or embryo level. For example, Jackson and Blundell (1963) reported that excised embryos of rugosa rose grown in Wales were non-dormant, whereas Svejda (1972) and Julin-Tegelman (1983), working with lots grown in Canada and Sweden, reported that excised embryos of this species required 3 to 4 weeks of chilling to become germinable.
Another source of satiation in dormancy status for rose achenes is a consequence of the post-maturation environment. Semeniuk and Stewart (1960, 1966) showed for several species that achenes from hips that had overwintered on the bush were more dormant than achenes from those same bushes collected and tested in the fall or stored dry and tested along with the field-overwintered achenes. This effect has also been noted by other workers (Jackson and Blundell 1963; Roberts and Shardlow 1979). It is probably best to collect rose hips soon after they reach maturity and to clean the collections immediately if seed dormancy status is an issue.
Because of the wide variation in dormancy-breaking requirements within each species, quality evaluations of rose seeds are usually carried out using tetrazolium staining (Gill and Pogge 1974). The achenes are first soaked in water for 24 hours. Firm pressure with a knife on the suture or a tap with a small hammer is used to split open the pericarp. The testa is then scratched or clipped at the cotyledon end and the seed is immersed in 1% tetrazolium chloride for 6 hours at room temperature. The testa is slit along the side and the embryo, which fills the seed cavity, is squeezed or teased out for evaluation (Belcher 1985). The excised embryo method may also be used, although it has little advantage over tetrazolium staining (Gill and Pogge 1974). For purposes of determining fill and chalcid infestation levels, x-radiography is suitable (Belcher 1985).
The preferred method in official testing is also tetrazolium staining (ISTA 1993), although stratification for 28 days at 3 to 5 °C is suggested for multiflora rose (AOSA 1993). For other rose species, the international rules (ISTA 1993) suggest an alternate method of 12 months of stratification, followed by germination in sand at 20 °C for 70 days. Germination is epigeal (figure 4).