Agricultural Biodiversity Weblog

In a recent post here I suggested that, despite frequent recourse to the comparison, genebanks are in fact not much like libraries, at least when it comes to deciding which of their contents can safely be jettisoned, or moved elsewhere. For this, librarians sometimes use the ejection criteria summarized by the acronym MUSTY: Misleading, Ugly, Superseded, Trivial and Your collection ¹. But those are not really all that appropriate for genebanks, I argued in the earlier post.

Well, if not MUSTY, what?

Genebanks should of course have a policy for deciding what to keep, but the only published example I know of is that of the international genebanks of CGIAR, which can be find in the Guidance Note for CGIAR Genebanks on Improving Accession Management.

The note points out that maintaining accessions is expensive, and needs to be rationalized, but that…

…improved understanding of diversity now provides an opportunity to reconsider and improve the composition and curation of a collection, by identifying potentially similar or redundant accessions that could be removed and gaps in collections that could be filled to increase or better cover inherent diversity for the crops in question. This would result in germplasm collections that will better address the goal of the conservation and use of crop germplasm and the changing needs of the stakeholders, clients and users within a global system. Having alternative options for curation or retention for specific classes of accessions provides options for rationalization and increases the transparency of curation decisions for the providers and users of the germplasm.

So CGIAR genebanks use four classes of curation: Fully curated, Partially curated, Archived, and Historical. Each represents a somewhat lower level of management activity, and therefore investment of time and money. And what accessions might be candidates for moving, say, to archived status? That would mean they “are believed to be alive and are stored in the genebank under optimal conditions for long-term survival, but without monitoring or distribution, while a final decision is made on their future so that, depending on their longevity, they can still be brought back to the curated collection or donated to other collections.” According to the note, lower priority accessions that might be candidates for archiving include:

• Accessions that are considered genetically similar to other accessions.
• Accessions from the same collection site and timeframe that are genetically similar and not the result of dividing a mixed original sample into multiple distinct accessions.
• Accessions that are outside the collection’s mandate and are best managed by others.
• Accessions that, based on a justified process for prioritizing accessions for conservation, are not considered for long-term conservation as part of the crop genepool, for example, an accession may come from a part of the genepool that is considered to be over-represented in the collection relative to other parts of the genepool.
• Mixed accessions that are no longer true to the original or have identity problems from physical errors and contamination.
• Problematic accessions that are beyond the ability of the genebank to continue their maintenance.
• Accessions of unknown identity or origin and have no historical records.

So, not so much MUSTY, as, what, SCRAPPY maybe?

S – Similar to other accessions genetically
C – Co-collected (same site/time), genetically similar
R – Rogue (outside the mandate)
A – Abundant in over-represented parts of the genepool
P – Polluted (mixed, contaminated, or identity problems)
P – Problematic to maintain
Y – Yesteryear’s mystery (unknown identity/origin)

I hope it catches on.

A really interesting recent episode of the 99% Invisible podcast alerted me to the fact that libraries occasionally get rid of books, something they call “weeding.” Now, genebanks are often compared to libraries, so I was interested to learn about the criteria librarians use in deciding what to weed. It turns out one popular — though not necessarily easy to follow — set of rules goes under the totally appropriate acronym of MUSTY:

M – Misleading, inaccurate, out of date. Unless you’re an official depository for books containing scientific theories that have since been disproved, don’t feel guilty about discarding books about NASA from 1975.

U – Ugly. Books ought to be beautiful, if at all possible. Books that are attractive will appeal to readers.

S – Superseded. If a better book comes along, don’t feel obligated to keep a former edition or favorite unless you are sure it has lingering value.

T – Trivial. People know I like books, and with the best of intentions they sometimes give me volumes that I really have no use for. Remember their thoughtfulness, thank them sincerely, but find a better home for those books – you will all be better off.

Y – Your collection: This book is no longer appropriate for your current passion. If you are finished learning everything there is to know about raising orchids and have moved on to quilting, donate the orchid books to a local club who can use them before they grow misleading, inaccurate and out of date for anyone else. Consider whether you have read it already and intend to do so again; or, if you haven’t read it, will you? Really and honestly?

So if a genebank found itself having to “weed” its collection, would any of these tests help at all? How does the genebank-as-library metaphor stand up to stress-testing?

Not hugely well, it seems to me.

Start with M. It’s hard to see how to apply this to germplasm. An accession might have inaccurate data associated with it, but nobody would get rid of it for that reason. It might be out of date in the sense of running down in viability, but that’s a reason for regenerating it, not binning it.

As for U, I guess this could refer to an accession that has been evaluated for certain traits and found wanting. But you never know what will happen with the next trait you evaluate for. And standards of beauty change.

Can a genebank accession be S for superseded? ² This might be trickier. Do you need to keep a really old batch of seeds after regeneration, say? But I don’t think you can really easily apply the concept at the accession level.

But maybe you can so apply T. Think of a plant originally collected as a potential forage that turns out not to be useful in that capacity at all. Does it still need to be part of a forage collection?

And that might also go for Y. Some genebanks do have themes: medicinals, forages, biosaline agriculture… If the mandate of the genebank changes, some stuff may need to find other homes.

MUSTY is not the only way that librarians use to guide their weeding. But to the extent that it is, it does not look like it translates to genebanks as easily as the prevalence of the metaphor might lead us to suppose.