Frequently, in studies on seed germination ecology, mistakes are made that lead to problems in determining the kind of dormancy and in extrapolating data to the field situation. In this paper, we discuss 22 of these issues and offer suggestions on how to avoid them.

The long-term utility of DNA banks is predicated on the stability of DNA during storage. The quality and yield of DNA extracted from seeds from four garden species, which varied in age from 1 to 135 years old, was used to examine the early stages of DNA degradation. Seeds that were 70 years old yielded high molecular weight DNA, which permitted amplification of a 650 bp segment of the internal transcribed spacer (ITS) region of the nuclear ribosomal DNA and a 487 bp segment of the plastid-encoded large subunit of rubisco (rbcL). DNA extracted from seeds estimated to be about 135 years old was degraded, but still permitted amplification when routine extraction procedures were used. The rbcL sequences obtained from c. 135-year-old cucurbit seed DNA were consistent with species identifications based on seed morphological features; however, ITS sequences from the same samples were determined to be of fungal origin. A comparison of our results with studies of DNA stability in leaf specimens suggested that DNA degraded within both tissues, but perhaps more slowly within seeds. Evidence that high-quality DNA can be extracted from old, non-viable seeds expands the utility of seed banks in preserving genetic resources.

Seed remains of common millet (Panicum miliaceum L.) were excavated from sites of ad 4th-century Darhan (Mongolia), and ad 15th-century Budapest (Hungary). Because the 15th-century medieval grains looked so intact, a germination test was carried out under aseptic conditions, which resulted in swelling of the grains but no cell proliferation or germination. Ancient DNA (aDNA) was extracted from the aseptic grains; analysed for amplified fragment length polymorphisms (AFLP), simple sequence repeats (SSR) and mitochondrial DNA (mtDNA); and compared with the modern millet cultivar ‘Topaz’. AFLP analysis revealed that extensive DNA degradation had occurred in the 4th-century ancient millet, resulting in only 2 (1.2%) AFLP fragments (98.8% degradation) amplified by MseCAA–EcoAGT, compared to the 15th-century medieval millet, with 158 (40%) fragments (60% degradation), and modern millet cultivar ‘Topaz’ with 264 fragments (100%). EcoAGT–MseCAA was found to be the most effective selective-primer combination for the analysis of medieval and modern millet. Eight AFLP fragments were sequenced after re-amplification and cloning. Microsatellite (SSR) analysis at the nuclear gln4, sh1, rps28 and rps15 loci revealed one SNP (single nucleotide polymorphism) at the 29th position (A→G) of rps28 locus, compared to modern millet. An mtDNA fragment (MboI), amplified at the 18S–5S ribosomal DNA (rDNA) locus in the medieval millet, showed no molecular changes compared to modern millet. The results underline the significance of aDNA extraction and analysis of excavated seeds for comparative analysis and molecular reconstruction of ancient and extinct plant genotypes.

Amaranthus tuberculatus is a weed species that has shifted emergence patterns over the past few years, presumably due to changes in seed dormancy in response to selection in agricultural fields. Although it is recognized that the seed dormancy phenotype is greatly affected by the environment, it is also acknowledged that the genotype plays a significant role. However, the importance of the genotype in determining intra-population seed dormancy variability, and the effect on emergence patterns, is not well understood. The objective of the present study was to determine the importance of the genotype on deep dormancy and the stratification-mediated dormancy alleviation in A. tuberculatus. Wild populations differing in seed dormancy were crossed and F2 families were generated. These families were used to determine narrow sense heritability of dormancy and stratification-mediated dormancy alleviation at the individual (hi2) and family (hf2) levels. hi2 ranged from 0.13 to 0.4 and 0.04 to 0.06 for the dormancy and stratification response, respectively. In the case of hf2, the values ranged from 0.76 to 0.91 for deep dormancy and from 0.33 to 0.58 for the stratification response. The genetic correlation between these two traits was below 0.075, indicating that different genes control them. High temperature strengthened the dormancy of deeply dormant seeds, making them less sensitive to stratification. However, high temperature promoted the germination of non-deeply dormant seeds. It is proposed that delayed weed emergence can be generated by selecting genes that control stratification response, and not necessarily only the genes that are directly responsible for deep dormancy.

The osmotin gene was introduced into tea [Camellia sinensis (L.) O. Kuntze] somatic embryos at the globular stage of development via microprojectile bombardment, to determine the effects of this stress-tolerance-related gene on the accumulation of storage reserves and acquisition of desiccation tolerance during embryo maturation. Changes in total soluble sugars, starch and proteins in response to osmotin were estimated in the stably transformed embryos, using standard biochemical and histochemical methods, after the globular embryos had matured to the heart stage of development. These changes also were compared with two types of controls, comprising (1) untransformed embryos and (2) embryos that were bombarded without osmotin DNA. The control embryos showed poor reserve accumulation and negligible germination but, in contrast, the embryos bombarded with the osmotin gene showed a several-fold increase in storage reserves, normal maturation and 57–63% germination. When these transformed embryos were desiccated to 15.5% moisture content, only 40% germinated, although a further increase in the accumulation of storage reserves occurred. In contrast, the control embryos became necrotic and failed to germinate. This study demonstrates the use of somatic embryogenesis as a tool for understanding seed recalcitrance in tea, and suggests that transformation with stress-tolerance genes can overcome the developmental problems associated with recalcitrant seeds.

Few data are available and little is known about spatial variation in post-dispersal seed removal at different levels throughout the geographic range of a plant species. Here, we compare post-dispersal seed removal by rodents within and among sites in two distinct regions, the south-eastern Spanish Mediterranean highlands and the Swiss Jura. Seed removal was assessed experimentally for four fleshy-fruited species (Berberis vulgaris, Crataegus monogyna, Rosa spp. and Taxus baccata) in heterogeneous sites with distinct microhabitats. A factorial general linear model (GLM) analysis was used to evaluate the relative influence of species-specific seed traits on removal at three spatial levels (microhabitat, site and region). In both highland ecosystems, live trapping revealed that rodents were the main seed removers: Apodemus sylvaticus was observed in both regions, whereas A. flavicollis and Clethrionomys glareolus occurred only at the Swiss sites. There was a significant difference in seed removal between regions, and the preferences of the rodents varied among plant species and microhabitats. Variation in rodent presence explained some differences in seed removal between regions and among microhabitats. Finally, the effect of rodent presence on seed removal differs due to both regional and microhabitat effects.

The ripening rice (Oryza sativa L.) caryopsis contains several maternal and embryonic tissues that transport assimilates along the partitioning pathway. Experimental access to transport steps in the pathway is limited by the separability of the tissues. Hence, the extent to which tissues can be mechanically and enzymatically separated was assessed. The caryopsis coat was isolated during mid-ripening, dissected and microscopically characterized. In centripetal order, the pericarp (epidermis, parenchyma-cell layer, cross-cell layer, tube-cell layer), inner integument, nucellar epidermis, aleurone layer and subaleurone layer adhered to each other. There was mechanical separation at the tube-cell layer, but not at the maternal/embryonic interface. Aleurone- and subaleurone-layer cells exposed on the inner surface of isolated caryopsis coats were macerated with Pectolyase Y-23 and manually sheared, which freed the endosperm and exposed the nucellar epidermis. Yield of endosperm cells increased linearly with the number of coats and reached a maximum after 1 h. The proportion of cells that were viable was approximately 10%. Subaleurone-layer cells contributed less than 25% to the yield. These results suggest that pectin polymers are necessary for adherence between the nucellar epidermis and the aleurone layer, and that enzymatic maceration is useful for separating the tissues at the maternal/embryonic interface.

The intact germination unit [seed+endocarp, (‘seed’)] of the two species of the Australian endemic genus Stylobasium (Surianaceae, Fabales) is water impermeable. Thus, the seed does not imbibe water or germinate until the integrity of the endocarp, which contains an outer layer of palisade cells, is disrupted. In this study, intact seeds of S. australe and of S. spathulatum, incubated on a moist substrate (for up to 2 years), did not imbibe water, whereas those that were mechanically scarified (both species), or dipped in boiling water for 30 s (S. spathulatum), imbibed and germinated. Thus, seeds of Stylobasium species have physical dormancy (PY). This is the first report of PY in Surianaceae, making it only the sixteenth family of angiosperms (no gymnosperms) in which this class of seed dormancy has been identified.