Setting acceptable seed viability thresholds for the World Vegetable Genebank collections

Abstract A genebank aims to conserve and distribute genetic diversity by keeping germplasm alive. For seed genebanks, this means that monitoring the viability of seed lots is a core activity. An initial assessment of seed lot viability—usually through a germination test—is used to decide whether or not a seed lot should be put into genebank storage. Subsequent germination tests are used to determine whether a seed lot can continue to be stored and made available for distribution. Both decisions are based on a defined criterion, the viability threshold (VT), the minimum accepted percentage viability of a seed lot. Since regeneration is a costly and risky operation, it is important that VTs balance the need to ensure high viability with what is practical in terms of how many accessions need to be regenerated. Here, we considered whether VTs of 85%, 70%, or 50% should be applied for species conserved in the World Vegetable Genebank in Taiwan (Headquarters [HQ]) and Tanzania (Eastern and Southern Africa offices [ESA]). Our assessment was based on considering the maximum proportion of seed lots that are marked as requiring regeneration because they have not passed the VT. Specifically, we applied a maximum failure rate of 30%. This value was chosen to represent a manageable regeneration workload, rather than as a universal biological limit for all species. Based on this principle, the widely accepted 85% VT could be adopted for 22 of 42 species conserved at HQ and for 10 of 22 species at ESA. The 50% VT should be adopted for 14 species at HQ and eight at ESA. For species where lower VTs are adopted, it is important to investigate why high germination is not achieved.