Laboratory characteristics of seeds of 37 species (41 seed types) from the Strandveld Succulent Karoo were used to predict seed bank types according to a modified key of ). Five seed bank strategies were recognized for this vegetation type, i.e. two with transient and three with persistent seed bank strategies. Of the 37 species investigated, 32% (all perennial species) had transient seed bank strategies, while 68% had persistent seed bank strategies. Seed dispersal of these 37 species was mainly anemochorous, although antitelechoric elements such as myxospermy, hygrochasy, heterodiaspory and synaptospermy were found among these species. The seed bank alone will not be sufficient to restore the vegetation of damaged land in the Strandveld Succulent Karoo, since many of the dominant species in the vegetation do not produce persistent seed banks. Many of these species may, however, be dispersed by wind into revegetation areas from surrounding vegetation. Topsoil replacement, seeding and transplanting of selected species will be essential for the successful revegetation of mined areas in this part of Namaqualand.

Seed dormancy was analysed in Geraniaceae, testing 29,103 seeds of 35 species for seed-coat impermeability to water. Seeds of wild and cultivated species from a number of different climatic/ecological zones (northern Germany, the Canary Islands and South Africa) were investigated. Seeds from cultivated plants and natural populations had similar (±2%) percentages of water-impermeable seeds. The percentages of impermeable seeds at maturity were also mostly constant (±2%) for two different years. The species analysed contained from 0 to 100% impermeable seeds at maturity, and could be classified into three quantitative types of physical dormancy: four species with no impermeable seeds (PY0), 10 species with a maximum of 80% impermeable seeds (PY80), and 21 species with more than 80% impermeable seeds (PY100). Species with different life cycles and from very different habitats, e.g. Geranium pratense or Pelargonium capitatum, belonged to the PY100 group. However, all of the succulent shrubs investigated from the genus Pelargonium were PY0 species. Five years of dry storage at 20 ± 5°C decreased physical dormancy in a species-dependent fashion. Erodium cicutarium, for example, had 95% impermeable seeds at maturity; all of them became permeable during the dry storage and began to germinate immediately after placing them on a moist substrate. Physical dormancy of other species, e.g. Pelargonium zonale, Pelargonium vitifolium or Geranium pratense, was not affected by dry storage. Thus, physical dormancy is a diversely differentiated feature in Geraniaceae with regard to both percentages of impermeable seeds at maturity and maintenance of dormancy under particular conditions.

Most Pinus species are obligate seeders. Thus, knowledge of germination characteristics can help in the understanding, prediction and manipulation of the regeneration and dynamics of pine forests. Seven pine species with contrasting habitat preferences and different genetic pairwise distances are present in the Iberian Peninsula and the Canary Islands: P. halepensis, P. nigra, P. pinaster, P. pinea, P. sylvestris, P. uncinata and P. canariensis. These seven pine species comprise an exceptional experimental set to test some questions related to germination traits, such as: (1) What are the effects of light and temperature on germination, taking into account interpopulation variability? (2) Is there any association of germination traits with habitat (montane versus lowland) preferences? and (3) What is the relationship between germination traits and the genetic distance between pine species? P. nigra, P. sylvestris and P. uncinata seeds showed faster germination rates. Seeds of P. nigra and P. sylvestris reached high total germination percentages in every temperature and light treatment, suggesting an opportunistic germination strategy. Unlike montane pines, lowland pines did show significant effects of temperature on germination response: final germination was higher between 15°C and 20°C than at warmer and alternating temperatures. Relatively low temperatures associated with the winter rainy season would favour germination of most of these species. Nested models showed that population variability was the main source of variation in germination response. Thus, there is no phylogenetic control of the germination response and, surprisingly, germination traits were not related to habitat preferences. As a consequence, we believe that studies of the germination characteristics of a pine species should consider different populations.

Changes in embryo dormancy of capeweed [Arctotheca calendula (L.) Levyns.] seeds in response to temperature were investigated to determine the nature of seasonal dormancy cycles. Primary embryo dormancy persisted for 2–3 months after seed collection and was then rapidly relieved when seeds were maintained at temperatures simulating summer soil surface temperatures. Embryo dormancy was also rapidly relieved in seeds maintained at constant temperatures, indicating that a daily temperature fluctuation was not necessary for the relief of embryo dormancy in capeweed. Dormancy relief was maximal at 40°C. Secondary dormancy was induced when seeds were maintained at low temperatures and a water potential of –1.5 MPa, suggesting that the onset of winter may postpone germination until a subsequent autumn. These results indicate that the dormancy cycles observed in capeweed seeds maintained on the soil surface are probably driven by seasonal changes in soil temperature.

Ecophysiological aspects of seed germination were investigated in four Mediterranean geophytes of the genus Muscari (Liliaceae): M. comosum (tassel hyacinth), M. neglectum (common grape hyacinth), M. commutatum and M. weissii. Experiments were performed at constant temperatures in the dark and under temperature and light conditions simulating those prevailing in nature during November–January, i.e. well into the rainy season of the Mediterranean climate. In all species, no primary dormancy was revealed, and germination occurred in a rather narrow range of cool temperatures (optimum at 10 or 15°C) and at a remarkably slow rate; both germination characteristics seem to be associated with autumn/winter seed germination and seedling establishment. Such a postulated strategy is ecologically advantageous within an unpredictable rainfall regime, known to prevail during the start of the rainy period of the Mediterranean climate. This strategy may also explain the spread of germination of M. comosum seeds over two consecutive years, observed by Theophrastus. Far-red light, simulating light conditions under a dense canopy, resulted in only a slight delay of germination compared to dark controls. Diurnal white light, qualitatively simulating natural daylight, caused a significant decrease of the germination rate in all four species studied. Moreover, white light was found to suppress considerably final seed germination (photoinhibition) in M. weissii and M. neglectum; in the latter species, prolonged imbibition under white light also led to the induction of secondary dormancy.