For those interested in rugosa seeds:
This link is to a long article so I have copied a rather large section pertinent to this thread.
“The seeds of R. rugosa are dispersed by several vectors, with seawater and birds being the most important. Several pieces of anecdotal evidence suggest a strong capacity for dispersal by sea currents. Rosa rugosa is thought to have first arrived on the coast of Norway by means of seed dispersal with sea currents, and furthermore dispersed by the same means from southernmost Norway to Lofoten in northern Norway (c. 1300 km) within only 40 years (Fremstad 1997). Both entire hips and individual achenes can float for several weeks. Jessen (1958) conducted a series of buoyancy experiments on both hips and achenes of R. rugosa and other species, using fresh and salt (24‰ salinity) water, regularly stirred, and kept at room temperature. All hips were able to float initially. A maximal floating time (time until the last hip sank) of 26, 40 and 42 weeks was obtained with over-ripe hips from three coastal populations. By this time hips had started to disintegrate or had been reduced to a thin membrane around the achenes. This is considerably longer than hips of native British Rosa species (Praeger 1913; Jessen 1958), both authors reporting R. pimpinellifolia L. to be among the longest-floating species with floating times of 6–7 weeks. Buoyancy of individual achenes also is good. Jessen (1958) found achenes that had been floating within entire hips for 40 weeks, to continue to float for another 4 weeks, after which the experiment was terminated. In some samples half of the achenes had sunk after 3 weeks, whereas in others all were still floating after 6 weeks. In general, the subsequent sowing of the achenes under garden conditions showed no reduction in germination percentages after floating. No consistent differences in buoyancy in fresh and salt water were found, either for hips or achenes. The same was the case for germination percentage. The floating ability of achenes is unique among Rosa species, with the exception of R. palustris and R. sweginzowii Koehne, both having swamps as their natural habitat. Achenes of these species can float for a few days only, however (Jessen 1958). Buoyancy is due to the spongy mesocarp (see Morphology).
Birds either feed on achenes, entire hips or hip flesh. Bird species feeding on entire hips may ingest achenes, and these are likely either to be regurgitated or to pass the intestines unharmed. Bird species feeding on achenes, e.g. greenfinch (Carduelis chloris) (Hegi Fl. ed. 3, 4; Persson, undated), destroy most of the ingested seeds, but unharmed seeds may be excreted on occasion. Thrushes and pigeons ingest achenes with the hip flesh, and may be the principal bird seed dispersers. Leege (1937) described how song thrush (Turdus philomelos), mistle thrush (T. viscivorus) and redwing (T. iliacus) passing the East Frisian Islands (Germany) on autumn migration ingest and regurgitate R. rugosa achenes, 3–12 per vomit pellet, which are c. 30 × 6 mm, bright yellow. Faecal and vomit pellets may be eaten by the thrushes themselves (Snow & Snow 1988), or rodents may eat achenes in them (Leege 1937). Gulls (Larus spp., e.g. the herring gull L. argenteus) have been suggested as probable seed dispersal vectors as well (Lohmeyer & Sukopp 1992), but documentation is lacking. Smaller birds, such as tits and warblers, may pick small pieces of hip flesh (see Animal Feeders or Parasites), but are unlikely to disperse the achenes.
Seeds of R. rugosa may potentially be dispersed by the same birds as other species of the genus Rosa occurring in the same habitats. Achenes of R. canina are frequently dispersed by blackbird (Turdus merula) and fieldfare (T. pilaris) in the British Isles (Snow & Snow 1988). Less frequently, song thrush, mistle thrush, redwing and woodpigeon (Columba palumbus) ingest and disperse Rosa achenes in the British Isles (Snow & Snow 1988), and waxwing (Bombycilla garrulus) in Central Europe and Scandinavia (Dingler 1912; Heintze 1916). Heintze (1917–18, 1932) found germinable seeds of Rosa in vomit pellets of magpie (Pica pica), hooded crow (Corvus cornix), nutcracker (Nucifraga caryocatactes) and jay (Garrulus glandarius). A further number of bird species feeding on rose hips and probably dispersing seeds were mentioned by Levina (1957) in a review of Russian literature primarily, e.g. grey partridge (Perdix perdix), raven (Corvus corax), jackdaw (Corvus monedula), starling (Sturnus vulgaris), and coal tit (Parus ater). In North America, pheasants (Phasianus colchicus and other species) have been reported to ingest and excrete viable seeds of Rosa spp. (Swank 1944). Common pheasant (Phasianus colchicus) is likely do the same in Europe, as it has been reported to feed on R. rugosa achenes during winter (Larsen 1944). It has been suggested that passage through the digestive tracts of birds or mammals may relieve mechanical (testa or pericarp induced) dormancy in seeds/fruits and enhance germination (see Traveset 1998). Krefting & Roe (1949) found that cold-stratified Rosa achenes germinated better after passage through grouse and pheasant guts.
Occasional short-distance seed dispersal is probably effected by microtine rodents harvesting and caching hips, and later eating the hip flesh, e.g. grey-sided vole (Clethrionomys rufocanus) (Ota & Jameson 1961). Other mammals reported to feed on rose hips and potentially disperse seeds are hare (Lepus europaeus), yellow-necked mouse (Apodemus flavicollis), bank vole (Clethrionomys glareolus), and red fox (Vulpes vulpes) (Levina 1957; Turček 1964).
(iv) Natural levels of recruitment
Vegetative spread by root-borne and stolon-borne suckers is the main form of propagation. No direct observations on the levels of recruitment from seed under natural conditions are available.
(d) viability and germination
Most achenes in a hip are usually filled and viable (Fagerlind 1948), with viability ranging from 84% to 94% (Svejda & Poapst 1972). However, Junttila (1974) found that achenes with a mass below a threshold of 4.5–5.0 mg, which is significantly below average, exhibited much reduced germination (7%).
Rosa rugosa has a relatively deep physiological seed dormancy and, in addition, some physical dormancy (according to the classification by Baskin & Baskin 1998). This is a common feature of the genus Rosa, but it is less pronounced in R. rugosa than in native British species (Tincker 1935). The seed dormancy is broken by a relatively long period of cold stratification, at least 5 weeks, but several months in order to obtain appreciable germination percentages. Dry storage (after-ripening) prior to stratification decreases the length of the chilling period required. When fresh achenes fall to the ground in the autumn, at least some will germinate in the following spring. Hip flesh contains germination inhibiting substances that can penetrate the achene pericarp. Thus, achenes that have dried out within the hip are more strongly dormant than fresh achenes. No difference between germination in light and darkness has been reported.
(i) Dormancy breaking
Several studies have investigated the effect of various pre-treatments and treatments on germination. Tincker (1935) found that some achenes could germinate without any pre-treatment, but that cold stratification at −2 °C to 2 °C, and various other stratification procedures, produced higher germination rates. Achenes of R. rugosa can germinate, albeit slowly, at very low temperatures, e.g. complete germination (maximal percentage achieved) of 50–80% after 7–11 months at 4 °C constant temperature (Crocker & Barton 1931; Rohmeder 1951; Junttila 1974; Tillberg 1983). The influence of the length of the stratification period on germination percentages has been investigated by Rohmeder (1951), Svejda & Poapst (1972) and Tillberg (1983). Rohmeder (1951) pre-treated achenes at 4 °C for periods up to 14 months, and subsequently let them germinate at either 20 °C or 25 °C. Without chilling, seed germination was barely detectable. With 4 months of chilling, maximum germination (which was only 20%) was reached after 2 months. There was little difference in germination rate and percentage between the 8- and 14-month chilling treatments. The maximum germination percentage (59%) was obtained with 14 months of chilling and 23 months at 20 °C. Germination percentage was consistently lower in the 25 °C treatment than in the 20 °C treatment. Svejda & Poapst (1972) found that pre-treatment of 20 weeks as compared to 4 weeks at 4 °C, and subsequent germination at 20 °C, gave increased germination percentage, but only up to 15% germination. Tillberg (1983) compared 11 chilling periods of 0–15 weeks at 4 °C, followed by 11 days at 20 °C in darkness. She found that chilling periods shorter than 5 weeks resulted in no germination, whereas chilling periods of 5 weeks or longer resulted in some germination, increasing to 30% after 15 weeks of chilling. Nyholm (1955) investigated the effect of germination temperature, and length of chilling period, on germination rate of seeds sown directly after harvest, or dry stored for 2 or 6 months. Germination percentage was consistently higher at 5 °C than at 1 °C or 10 °C, when seeds were kept at constant temperature. At 15 °C and 20 °C no seeds germinated. Transfer of seeds to germination at 20 °C after variable chilling periods showed that after-ripening takes 3 to 4 months. Dry storage of seeds resulted in lower germination percentages.
Repeated cycles of cold and warm treatment seem to induce germination to higher percentages. Junttila (1974) subjected 78 Fennoscandian samples of achenes to cycles of 8 weeks at 4 °C and 4–6 weeks at 20 °C/15 °C (day/night), and found germination percentages after one, two and three cycles to be 23.3%, 46.0% and 53.6%, respectively. Similarly, Svejda & Poapst (1972) subjected achenes to 20 weeks at 4 °C, 8 weeks at 20 °C and 12 weeks at 4 °C, before final germination at 20 °C, and found germination percentage to increase from 43% to 76% after the second chilling.
The effect of hip flesh on dormancy was investigated by Crocker & Barton (1931). They compared the final germination percentage of fresh seeds and seeds dried within hips on the shrub, either sown immediately or after 3 years of dry storage. Seeds allowed to dry on the plant had lower germination percentage (27%) than had green seeds (50%). Prolonged dry storage had little effect on the germination percentage of green seeds, but resulted in increased germination percentage in seeds dried within hips. However, inhibiting substances are leachable (Jackson & Blundell 1963), and probably soil water can leach them. Similarly, the effect of germination inhibitors exuded by the pericarp and testa can be counteracted by leaching or germination in the presence of charcoal. Leaching of achenes in distilled water for 24 h prior to cold stratification increased germination percentages (Svejda & Poapst 1972). However, leaching had no effect on germination of achenes where the first germination treatment at 20 °C for 4 or 8 weeks was followed by a new 12-week chilling period, before the final germination treatment. Thus, the effect of inhibitors exuded from pericarp or testa seems to diminish with sufficiently long cold stratification. Cold stratified achenes germinating in the presence of charcoal have been shown to increase germination percentage almost twofold, presumably because the charcoal absorbs germination-inhibiting substances from the imbibition water (Julin-Tegelman 1983).
(ii) Physiological dormancy
Several authors have investigated the possible mechanisms of seed dormancy. Jackson & Blundell (1963) demonstrated that the pericarp, and to a less degree the testa, exert strong inhibition on germination of ripe embryos. They found that naked embryos excised from achenes stratified at 2 °C for 3 months germinated to 100% within 3 days at 26 °C. Seeds (embryo and testa) without pericarp germinated to a maximum of 45% within the same time, whereas entire achenes germinated to a maximum of 14% after 2 weeks. They identified the substance responsible for inhibition as indole-3-acetic acid in leachate from entire achenes. In addition, they showed that exogenously applied cytokinins (purine and gibberellic acid) overcame inhibition and induced dormancy release in R. arvensis. Svejda & Poapst (1972) showed that the inhibitory substances diminished with cold stratification (4 °C), and that warm stratification had no effect unless applied prior to cold stratification, in which case it shortened the chilling period required and increased germination percentages. Julin-Tegelman (1983) and Tillberg (1983) completed the studies of Jackson & Blundell (1963) by showing that even excised embryos need at least 3 weeks of chilling (at 4 °C) to show some germination (5 weeks chilling for 100% germination to appear), and that control achenes kept at 17 °C showed no germination.
Julin-Tegelman (1983) and Tillberg (1983, 1984) presented studies of endogenous inhibitors and promoters during dormancy release and germination on the same seed material (R. rugosa var. rubra). The concentration of free and bound cytokinin-like substances in achenes shows marked, but transient, peaks during cold stratification, as compared to non-chilled achenes (Julin-Tegelman 1983). Bound cytokinin-like substances also peaked during the first few days of germination. Zeatin and ribosylzeatin were identified tentatively as the major constituents in free and bound cytokinin, respectively. Dormant, unimbibed achenes contain exceptionally high concentrations of abscisic acid (2.7 µg g−1 fresh unimbibed weight), a compound often suggested to be a germination inhibitor, and this concentration decreased strikingly during the first 2 days of cold stratification (4 °C). However, the decrease in abscisic acid concentration was even bigger in achenes imbibed and kept at 17 °C (Tillberg 1983). Concentration of indole-3-acetic acid, the most important auxin, decreased rapidly during cold stratification, and remained at low levels, but showed a similar pattern as abscisic acid (Tillberg 1984). During germination the indole-3-acetic acid concentration fluctuated, but it appeared low when germinability was increasing. A peak in indole-3-acetic acid concentration after 2 days of germination was preceded by an increase in the concentration of cytokinin-like substances, so probably cytokinin levels affect auxin levels in germinating R. rugosa achenes.
To sum up, the three studies show that abscisic acid does not play any direct inhibitory role in either dormancy release or germination, whereas cytokinin-like substances appear to play a promotive role in both dormancy release and germination, and indole-3-acetic acid in germination only. A precursor of indole-3-acetic acid, N-malonyl-D-tryptophan, which has been shown to occur in seeds and seedlings of some plant species, was absent from R. rugosa seeds and seedlings, and detectable as traces in wilted leaves only (Gamburg et al. 1991).
(iii) Physical dormancy
The pericarp seems to exert a mechanical inhibition on germination of the embryo, in addition to the physiological inhibition described above. Although the pericarp is somewhat water-permeable, it severely restricts the penetration of water into the embryo. Svejda (1972) showed that the weight gain during imbibition was 43% for newly harvested entire achenes, but 112% for cut achenes. It was also shown by staining, that the weight gain in entire achenes was mainly located in the pericarp and testa. A similar result was obtained by Julin-Tegelman (1983) for unstratified dry achenes and excised embryos. For entire achenes the weight increased 43% in the first day and then levelled off at 55%. For excised embryos the weight increased 20% in the first day and then levelled off at 30%. However, mechanical and chemical scarification does not necessarily result in higher germination rates. Scarification in 70% sulphuric acid for 3.5 h resulted in germination of achenes previously chilled at 4 °C for 5–9 weeks, whereas unscarified achenes barely germinated (Julin-Tegelman 1983). By 11 weeks of chilling, however, the difference between scarified and unscarified achenes was negligible. Mechanical scarification by vigorously shaking dormant achenes with coarse sand for 2 h prior to cold stratification was attempted by Tillberg (1983), but did not increase germination rate. Scarification by sand would otherwise seem to be a plausible mechanism in nature, where seeds are dispersed by sea currents and washed upon sandy beaches.”
http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2745.2005.01002.x/full