The conservation of crop genetic resources, including their wild relatives, is of utmost importance for the future of mankind. Most crops produce orthodox seeds and can, therefore, be stored in seed genebanks. However, this is not an option for crops and species that produce recalcitrant (non-storable) seeds such as cacao, coffee and avocado, for crops that do not produce seeds at all; therefore, they are inevitably vegetatively propagated such as bananas, or crops that are predominantly clonally propagated as their seeds are not true to type, such as potato, cassava and many fruit trees. Field, in vitro and cryopreserved collections provide an alternative in such cases. In this paper, an overview is given on how to manage and setup a field, in vitro and cryopreserved collections, as well as advantages and associated problems taking into account the practical, financial and safety issues in the long-term. In addition, the need for identification of unique accessions and elimination of duplicates is discussed. The different conservation methods are illustrated with practical examples and experiences from national and international genebanks. Finally, the importance of establishing safe and long-term conservation methods and associated backup possibilities is highlighted in the frame of the global COVID-19 pandemic.

The ability of crop seeds to retain their viability over extended periods of uncontrolled temperature and/or relative humidity conditions has not been widely investigated, although this is an important issue for genebank management. We report here the response of 18 crop species to storage for up to 26 years at 20.3 ± 2.3°C and 50.5 ± 6.3% relative humidity. Germination rates decreased in a sigmoid fashion, but the curve parameters were species characteristic. Pea, common bean and maize seeds retained their viability over the longest period (23, 21 and 19 years, respectively). In contrast, chive seeds survived for only 5 years and lettuce for 7 years. In addition to this interspecific variability, there were also indices for intraspecific variability, particularly in bean and chive seeds, just as in collard, lupin, poppy, wheat and maize seeds. A significant correlation was obtained between germination performance in the laboratory and seedling emergence following autumn sowing. Seeds in which oil was the major seed storage component were more short lived, whereas carbohydrates or proteins did not show an effect on seed longevity.

Wheat sheds tricellular short-lived pollen at maturity. The identification of viable pollen required for high seed set is important for breeders and conservators. The present study aims to evaluate and improve pollen viability tests and to identify factors influencing viability of pollen. In fresh wheat pollen, sucrose was the most abundant soluble sugar (90%). Raffinose was present in minor amounts. However, the analyses of pollen tube growth on 112 liquid and 45 solid media revealed that solid medium with 594 mM raffinose, 0.81 mM H₃BO₃, 2.04 mM CaCl₂ at pH5.8 showed highest pollen germination. Partly or complete substitution of raffinose by sucrose, maltose, or sorbitol reduced in vitro germination of the pollen assuming a higher metabolic efficiency or antioxidant activity of raffinose. In vitro pollen germination varied between 26 lines (P < 0.001); between winter (15.3 ± 8.5%) and spring types (30.2 ± 13.3%) and was highest for the spring wheat TRI 2443 (50.1 ± 20.0%). Alexander staining failed to discriminate between viable, fresh pollen, and non-viable pollen inactivated by ambient storage for >60 min. Viability of fresh wheat pollen assessed by fluorescein diacetate (FDA) staining and impedance flow (IF) cytometry was 79.2 ± 4.2% and 88.1 ± 2.7%, respectively; and, when non-viable, stored pollen was additionally tested, it correlated at r = 0.54 (P < 0.05) and r = 0.67 (P < 0.001) with in vitro germination, respectively. When fresh pollen was used to assess the pollen viability of 19 wheat, 25 rye, 11 barley, and 4 maize lines, correlations were absent and in vitro germination was lower for rye (11.7 ± 8.5%), barley (6.8 ± 4.3%), and maize (2.1 ± 1.8%) pollen compared to wheat. Concluding, FDA staining and IF cytometry are used for a range of pollen species, whereas media for in vitro pollen germination require specific adaptations; in wheat, a solid medium with raffinose was chosen. On adapted media, the pollen tube growth can be exactly analyzed whereas results achieved by FDA staining and IF cytometry are higher and may overestimate pollen tube growth. Hence, as the exact viability and fertilization potential of a larger pollen batch remains elusive, a combination of pollen viability tests may provide reasonable indications of the ability of pollen to germinate and grow.

The accumulation of reactive oxygen species has been associated with a loss of seed viability. Therefore, we have investigated the germination ability of a range of seed stocks, including two wheat collections and one barley collection that had been dry‐aged for 5–40 years. Metabolite profiling analysis revealed that the accumulation of glycerol was negatively correlated with the ability to germinate in all seed sets. Furthermore, lipid degradation products such as glycerol phosphates and galactose were accumulated in some seed sets. A quantitative analysis of nonoxidized and oxidized lipids was performed in the wheat seed set that showed the greatest variation in germination. This analysis revealed that the levels of fully acylated and nonoxidized storage lipids like triacylglycerols and structural lipids like phospho‐ and galactolipids were decreasing. Moreover, the abundance of oxidized variants and hydrolysed products such as mono‐/diacylglycerols, lysophospholipids, and fatty acids accumulated as viability decreased. The proportional formation of oxidized and nonoxidized fatty acids provides evidence for an enzymatic hydrolysis of specifically oxidized lipids in dry seeds. The results link reactive oxygen species with lipid oxidation, structural damage, and death in long‐term aged seeds.