Evolutionary diversity in the Millennium Seed Bank

John Dickie, Head of Seed and Lab-based Collections at Kew, describes an approach to measuring and maximising the evolutionary diversity conserved in seed banks of wild species.

Seeds in sub-zero storage facilities at Millennium Seed Bank

The Millennium Seed Bank – conserving plant diversity

Since the middle of the last century, crop gene banks have sought to mitigate the loss of potentially useful diversity in older varieties and wild relatives through the mass storage of seeds. This relies on the capacity of the seeds of the majority of seed-plant species to survive for potentially many hundreds of years, air-dried and deep frozen. The Royal Botanic Gardens, Kew was an early adopter of this advantageous physiology, applying it to wild species and setting up a seed bank at its satellite garden, Wakehurst Place in West Sussex.

That initiative has since grown into the Millennium Seed Bank Partnership, a worldwide network of botanic gardens and others conserving wild plant biodiversity as seeds in long-term storage. The Global Strategy for Plant Conservation acknowledges the vital complementary role of ex situ conservation, including seed banks and living collections in botanic gardens, for plant species conservation worldwide.

During its first ten years, the Millennium Seed Bank Project achieved its target of conserving 10% of the world’s seed-bearing flora. Now, the MSB Partnership is on target to have 25% of those species in safe storage by 2020. It is clearly necessary to prioritise the seed collecting activity and up to now efforts have been focused substantially but not exclusively on threatened species, those with restricted distributions, and those non-domesticated species that are useful to man - the ‘Three E’s’ strategy: endangered, endemic and ‘economic’. The strength of that strategy has been the practicability of its implementation and the priority given to species under greatest threat of either global or local extinction, together with those having most potential for sustainable use.

Phylogenetic diversity

However, in parallel with the growth of the MSBP, the potential importance of capturing phylogenetic diversity (PD) for in situ conservation has been given increasing attention, for both animals and plants. Phylogenetic diversity represents the total evolutionary history of a set of species, and thus its evolutionary potential. It is measured by summing the total phylogenetic distance, span and branch lengths of an evolutionary tree among a set of species.

Phylogenetic diversity doesn’t always correlate with species diversity and species counts alone can therefore be poor representations of evolutionary history and diversity. As a result of this, conservation scientists at the MSB have been keen to estimate just how much phylogenetic diversity is held in its collections (which were made without explicitly targeting PD), and to find out how to prioritise collecting so as to maximise this important component of diversity.

Rich Grenyer and his group at Oxford University are leaders in phylogenetic diversity research (hitherto in an in situ context) and the research described here came out of a collaboration between Kew and the Oxford group. As part of her MSc studies, Kate Griffiths led the analysis, supervised by Rich Grenyer. In addition to the MSB team of Sharon Balding, Tim Pearce and John Dickie, Kew Herbarium’s legume expert Gwil Lewis contributed valuable access to the latest legume taxonomy and phylogeny. 

Assessing the phylogenetic diversity of the complete MSB collection was not practical, so attention was focused on the Leguminosae (Fabaceae), because of both their strong representation in the MSB’s collections and their global economic importance. The family is well studied and construction of a genus-level phylogeny for the whole family was a prerequisite for the analysis. 43% of legume genera are represented in the MSB’s collections and 18% of species. Characteristics used in the analysis included: collection location; threat level; degree of endemism and economic value; compiled from sources that included Kew’s SEPASAL database, the IUCN Red List and the Sampled Red List Index. 

The results – opportunities for enhancing evolutionary diversity in the MSB’s collections

The analysis confirmed that the MSB’s ‘3 E’s’ targeting strategy has paid off, with the representation of genera and species in the bank satisfying one or more of those criteria being substantially higher than that across the entire legume phylogeny. Prioritising collection of the most important and at risk species is an appropriate focus for seed banking and the analysis revealed that aiming for them has also led to the capture of more than half (62%) of the total PD in the Leguminosae.

While it appears that focusing on endangered and useful species has not compromised overall PD, the results of the analysis do suggest that concentrating on endemic species has the potential to limit the PD of the MSB’s collections in future. This highlights a potential risk for region-focused collections, with a high degree of phylogenetic clustering among range-restricted legume genera apparent.

An important finding from the study is that large gains in PD could be made by targeting specific genera based on their ‘irreplaceability score’, which can be thought of as evolutionary distinctness in this context. The 62% PD representation among current MSB legume collections could be increased to 68% by the strategic addition of just 10 genera, to 87% with the addition of 100 new genera, and to 95% with 177 new genera. Such strategic additions are already being built into the plans for the MSB’s future legume collecting, in terms of both target species and their geographic locations.

Furthermore, it is also hoped to extend the analysis to several other important plant families, including grasses and palms, both economically important. The latter is particularly interesting, as it contains a high proportion of species (perhaps around 70%) with seeds that cannot be dried and thus cannot be stored in conventional seed banks (recalcitrant). It will be valuable to know how much this constraint restricts PD representation in ex situ collections of palm seeds.

Although the approach developed in the study is primarily aimed at seed banks, it could also be usefully applied to a broad range of ex situ conservation collections.


Forest, F., Grenyer, R., Rouget, M., Davies, T.J., Cowling, R.M., Balmford, A., Faith, D.P., Manning, J.C., Procheş, Ş., van der Bank, M., Reeves, G., Hedderson, T.A.J. & Savolainen, V. (2007). Preserving the evolutionary potential of floras in biodiversity hotspots. Nature 445: 757-760. Available online

Griffiths, K.E., Balding, S.T., Dickie, J.B., Lewis, G.P., Pearce, T.R. & Grenyer, R. (2014). Maximizing the phylogenetic diversity of seed banks. Conservation Biology Early View.  Available online