Agrobiodiversity education in context

A piece on “Generating the next generation” by Nigel Chaffey in his latest, always indispensable, Plant Cuttings had me trawling around for an hour or so last night amid botanical teaching resources, looking for stuff that might be relevant to agricultural biodiversity. It’s not a great haul, alas.

Teaching Tools in Plant Biology, published by the American Society of Plant Biologists, does have Genetic Improvements in Agriculture, but it’s behind a subscription wall. The American Society of Plant Biologists has pages of resources for K-12 and higher education, but the focus seems to be on biotechnology. Fortunately, the Plant Science TREE (Tool for Research Engaged Education), from the Gatsby Plant Science Summer Schools, does have a useful, freely available section on Plants and People.

I was also momentarily encouraged by seeing an old friend posing in his rice genebank on the homepage of Science & Plants for Schools website:

But the caption he is lumbered with is, weirdly, about the role of plant sciences in “developing cures for diseases.” And anyway nothing happens when you click on him. However, feeding the world is also mentioned (phew), and I was in the end able to find something on genebanks and plant breeding. I wouldn’t call the coverage comprehensive, though. Nor systematically presented.

There is, of course, a place for teaching resources specifically for agrobiodiversity, but one would like to see the subject a little better integrated into the wider plant sciences education universe. Wouldn’t one? Well, not if there are many students like Katie DeGroot.

A roadmap to better mapping

Geographers and cartographers often use 2-3 three different software packages for data analysis: they will probably never settle around one tool, online at that, and create a ‘community’ of users there. Instead, the NGOs interested in such a tool should rather offer geo-info advice and look at light open-source GIS software to distribute: how many development workers in the field have had difficulties with the (basic) tabular conversions associated with GPS data? Many many me thinks.

That’s Cédric Jeanneret-Grosjean on online mapping resources. What he’s saying is that they, er, should not be online. Bold. Very bold. But a model that has in fact been followed, at least for the spatial analysis of biodiversity, agricultural and otherwise. And with some success. Maybe time for the crop distribution modellers to try it?

Let’s remember this is important. We’re not just arguing about how to make prettier maps. Identifying what constraints are going to be most significant, when, where in the world, for each crop, is going to be crucial in setting breeding agendas for the next 20 years and more. Breeders need to be able to explore and interrogate these future suitability maps, and explain what they get out of them to their bosses and the policy-makers above them. It’s important to make them as accessible and easy to use as possible. What we have at the moment is not fit for purpose.

Digging up the early history of an early peanut

I recently learned in a throw-away comment during the Q&A after a talk about the Vavilov Institute (VIR) genebank that until just a few years ago a single accession was the main source of early maturity in peanut, a line called Chico. And that variety supposedly traced back to Russia, not a place I for one usually associate with groundnut cultivation, or indeed breeding. Worth a little digging (pun intended).

Let’s start with Chico’s significance. Our go-to guy for groundnut genetic resources confirmed that it was indeed one of the most important sources of early maturity, together with Gangapuri and JL 24. It was used extensively at ICRISAT and various other breeding program for many years, although a number of other sources have now been identified in the mini-core collection.

Next, where did it come from? It is clear in the registration notification 1 that it was in fact a selection from a line from a Russian breeding programme:

Chico was developed by line selection from PI 268661, an introduction into the United States in 1960 from Rhodesia. It had come originally from Krasnodar, USSR, where it was designated Arachis Line No. 370 from ‘VNIIMK 8459.’

PI 268661 is still available in GRIN. That links to the original entry in the plant introduction book for 1960, which reads like this:

268661. SB52. ‘Apaxuc 370’. From U.S.S.R.

This was part of a large consignment of peanuts from what was then Northern Rhodesia and is now Zambia.

268491 to 269135. ARACHIS HYPOGAEA L. Fabaceae. Peanut.
From Rhodesia. Seeds presented by the Mount Makulu Research Station, Chilanga. Received Oct. 11, 1960.
AB denotes alternate branching bunch variety.
AR denotes alternate branching runner variety.
SB denotes sequential branching bunch variety.
BC denotes from Tozi collection, Sudan.
IN59 denotes a 1959 introduction into Rhodesia.
SR denotes Southern Rhodesia strain collected in 1959.

Mount Makulu, incidentally, is still where the Zambian national genebank is housed.

So somehow or other a peanut variety called Apaxuc 370 from a place called Krasnodar, which was in some way derived from VNIIMK 8459, ended up in Northern Rhodesia and, along with many other peanuts, was in due course sent to USDA in 1960 by staff of Mount Makulu. Where it no doubt hung around for a while, but was eventually evaluated and identified as being interesting. A line was then selected and released in 1973 by the Georgia, Virginia and Oklahoma Agricultural Experimental Stations. And the rest is history.

But can we go any further back in time? I asked our friends at VIR and it seems not, unfortunately. “Krasnodar” is in fact the Research Institute of Oil Crops in Krasnodar (VNIIMK is its Russian acronym).

Their names very often consist of the abbreviation “VNIIMK” plus a breeding number. However, there are no accessions numbered “8459” in the collection. Varietal names include numbers of 4 digits, but they always begin with “1”.

Oh dear. What about that Apaxuc 370? First, “apaxuc” is clearly just a rendering of the Russian for peanut (арахис). From VIR again:

There are no lines numbered 370 in VIR’s collection as well. In our peanut catalogue No. 307 corresponds to the variety Stepnyak bred at VNIIMK and used for oil production purposes. It was registered with the catalogue in 1945. The pods of this variety are quite different from those of var. Chico in size and shape: they are larger and have a deep constriction between seed vessels.

Another dead end. VIR does have Chico in its genebank (catalogue No. 1199), but that came in 1980, from the US, with no further information about its history. Bit of a mystery, though, about how its progenitor got from Krasnodar to southern Africa.

There are no data in our documentation on any germplasm exchange with Rhodesia in the middle of the 20th century. Neither there were any additions to the collection from South African countries whether by collecting missions, seed requests or research visits.

And an even bigger mystery about where that progenitor came from in the first place. Was it something Vavilov himself collected on his South American travels, perhaps. I’d like to think so, but I fear we may never know. Not from the existing passport data, anyway. Maybe someone has done some molecular work, though?

Still waiting for a decent way to map change in climate suitability

There seems — inevitably — to be something of a competition out there to produce maps of changes in climatic suitability for different crops. And I’m not saying that’s necessarily a bad thing. The recent launch of FAO’s GAEZ Data Portal gives us a much-needed alternative to the all-but-unusable offering by CCAFS. Sad to say, it is not much of an alternative.

Here’s the quick version. CCAFS’s Adaptation and Mitigation Knowledge Network allows you to map current suitability, future suitability and change in suitability for a few crops. The example below is change of suitability for Phaseolus beans in Africa. I wasn’t able to get rid of those extraneous place markers, nor to export the results other than by screengrab, nor to import my own data to super-impose, nor to include the legend in any sensible way.

FAO tries to fill the gaping hole left by this well-meaning but flawed tool by providing something that: has menus which are extremely cumbersome to navigate; does not include change in suitability, but then doesn’t let you combine present and future suitability to do your own analysis; does allow you to choose different climate models, but again not to combine the results; does allow various download options, none of them particularly useful, and then only if you register; and, needless to say, doesn’t allow you to import data.

It’s hard to say how one could better replicate, and in fact accentuate, the bad features of one tool, while ignoring whatever good things it had. Frankly, I’m running out of patience with websites which are clearly designed by one set of geeks for another set of geeks. Meanwhile, users — you know, the people for whom this stuff is allegedly produced — are left to cry over their keyboards, and think of what might have been. And regret the fact that they’ll never get back the hour they’ve just wasted.