Is more always better? Of course not. The Western Farm Press ((West of where, you might well ask, largely in vain, so I’ll tell you: the US.)) had an intriguing headline a couple of days ago: “Genebanks increase in importance, number 1,700 worldwide”. The bulk of the article wasn’t really about the number of genebanks, but rather about the Crop Genebank Knowledge Base put together by the System-wide Genetic Resources Programme of the CGIAR. ((Strictly speaking it wasn’t so much about the CGKB as a copy of the SGRP’s launch statement, very slightly rejigged.)) The one thing that Western Farm Press did do was to conflate an increased number of genebanks with their increased importance. I found this particularly odd given a recent piece by Cary Fowler, Executive Director of the Global Crop Diversity Trust, which argued forcefully that 1700 genebanks was almost certainly about 1694 too many. Here’s a taste:
In the early 1970s, the grand old men and women associated with the International Board for Plant Genetic Resources suggested that the world needed 6-7 international genebanks to conserve crop diversity and supply the needs of plant breeders and researchers. This, presumably, was in addition to a few good genebanks already operating. Today, FAO’s registry of genebanks lists 1700 facilities! Were the experts off by a factor of 250? Or do we have a bubble?
It is tempting to equate more genebanks with more and better conservation. But in the case of genebanks, the establishment of more and more facilities has not automatically translated into expanded conservation because (a.) many of those facilities were stocked with samples already being stored elsewhere, and/ or (b.) professional and financial capacity were insufficient to ensure effective conservation over time.
The motivations for establishing so many facilities were certainly positive. Scientists and administrators realized crop diversity was endangered and wanted to save it. But, as political interest heightened, saving something important evolved into shielding it from others. Sharing between farmers, scientists and countries – the very act that lay at the heart of the establishment of the banks – morphed into a negative.
Today, the tiny “network” of genebanks to which the early experts referred provides the overwhelming majority of samples distributed to farmers and breeders. Very little escapes from other collections. Dozens of “important” genebanks have not conveyed samples even to domestic researchers, much less foreign ones in recent years.
Of course I tried to ask the Western Farm Press whether they really thought that the number of genebanks was a measure of their importance, rather than, as Fowler suggests, evidence that people are not taking them seriously. Unfortunately, the WFP website is badly broken, asking me to verify a word that doesn’t appear in at least three browsers on two different platforms. Honestly. Ho hum.
Frankly, it’d be nice if WFP mended its site, although it isn’t exactly a matter of life and death. It’d be much nicer if the global genebank community (for that is what it is, as Fowler reminds us) cleaned up its own act and moved towards the rationally organized system that future food security requires, which is a matter of life and death.
IBPGR (founded in 1974) had nothing to do with the recommendation by the CGIAR TAC in 1972 to establish or amplify a global network of genebanks. This was done at the famous Beltsville meeting 20-25 March 1972 which advised: “ten regional genetic resources centres; plus a smaller number of crop-specific centres – including the existing and proposed “International Centres” being supported by members of the Consultative Group.
At the same meeting they recommended a “Coordinating Committee – located in FAO and supported by a small central staff of three specialists”. This became IBPGR. The entire operation was to be funded for 5 years, closely in association with FAO, and there was to be a “trust fund” (sound familiar?).
I have since then worked at three of the 15 or so recommended genebanks. The were two problems a) two regional centres morphed into national genebanks and lost their regional focuscoverage (Izmir and Addis Ababa); b) other supposed regional centres simply carried on as national genebanks (Chapingo, Campinas and IARI in New Delhi).
But as countries were mainly interested in genebanks as a service to breeders (or as deposits of introduced crops – as for the USDA and the Kenya genebank) and as genebanks were often associated with crop-specific national institutes, the large number was not a problem, as it was closeness to national breeders that mattered and not secure storage and international exchange.
The problem of exchange in recent years has been first, the NGO `biopiracy’ campaign that scared the pants off many countries; second the CBD provisions for `national sovereignty’ which left the question open as to how countries were to be rewarded; and third, the FAO Seed Treaty, which offers very little to countries which think they are sitting on a genetic gold mine other that yet more on-farm conservation.
How to change this is the question. Rather than emphasize the value of national genebanks (often of questionable value to countries themselves and anyone else for smaller countries) there should a strong emphasis on the value of introduced crops (70% of production value for Latin America and Africa): “If you don’t distribute samples, you don’t get samples” (but this is made difficult by the fact that the excellent genebanks 0f USDA and the CGIAR actually do distribute samples unconditionally, and also advanced lines and varieties).
National genebanks may be broke because far too many were established, often as a quid pro quo for permission to collect samples in the heady 1970s and 80s.
In my review of CGIAR genetic resources in 1991 (the first?? use of `agrobiodiversity’) I concluded that: “For continued international access to well-managed collections, there is no foreseeable alternative to the present CGIAR system…” Twenty years on this still seems to be the case.
Apologies for the length but the Beltsville report was on my desk.
No apology necessary for setting the record straight, if that’s what you did. But a close textual reading of Cary Fowler’s piece does not, it seems to me, claim that IBPGR made the recommendation. It says that people associated with IBPGR made that recommendation. Nevertheless, thanks for the additional detail.
Jeremy – I’m pedantic: when the recommendation was made the grand old men and women were not associated with IBPGR, which did not then exist. We need a `later’ to be inserted between women and IBPGR.
“… yet more on-farm conservation …”
Are readers to infer from this a belief that there is somehow ‘too much’ on-farm conservation? Or simply not enough effective ex situ?
Isn’t well organised and resourced on-farm conservation vital?
Good question. We’ll have to wait to see what Dave Wood intended us to infer.
Jeremy: Let me take up the gauntlet.
It all depends what we mean by ‘on-farm conservation’. I have no problem whatever with the repatriation of lost varieties from genebanks or the enrichment of on-farm resources by the kind of projects done by John Witcombe and his co-workers, where varieties are brought in from afar and modern varieties such a PB 1 and IR 64 used as parents in crosses for delivery – with extensive uptake on farm.
I also like the spread of the large number of stem- and leaf-rust resistance genes in modern wheats, obtained from an assortment of wheat wild relatives. These genes are now being conserved all over the place in modern wheat varieties, as are the key dwarfing genes of wheat and rice in widely used Green Revolution varieties (as in the entire D-genome of wheat from Aegilops tauschii and genomes from any other wild species used in breeding a range of crops). These genes will now be conserved as long as we farm.
But this is open-system dynamic conservation, under farmer management, with varieties coming and going as the farmers wish, and with associated ex-situ conservation allowing breeders to incorporate a large number of key genes into varieties that would then be grown all over the place.
In contrast, there is closed-system on-farm conservation, supposedly dynamic as varieties evolve over time (but the evidence in still not there), and, with the idea ‘local is best’, no varieties or crops from elsewhere are thought to be necessary. I think this is technically highly dubious and also certainly unethical, as it prevents the flexibility offered to farmers by new – and often better – diversity. Unfortunately, this is what, for example, FAO will go for, as the groundwork was done long ago by an FAO staff member (Bennett) and then adopted by NGOs as part of their campaign against the Green Revolution.
Let’s look at the history. There was a FAO/IBP Technical Conference on Genetic resources in 1967. The results were edited by Bennett (1968, expanded by Frankel and Bennett in 1970 in their book ‘Genetic Resources in Plants – Their Exploration and Conservation’). Bennett argues that the purpose of conservation of what were then called ‘primitive cultigens’ is “… to conserve material so that it will continue to evolve.” Such changes were thought to be desirable as long as ‘genetic erosion does not take place’. The Frankel and Bennett (1970) paper goes further. It makes two emphatic arguments (p. 13): that in centres of crop diversity there is “a direct and pressing interest in the use of indigenous populations … since in many instances it has been found that some selections from indigenous crops have proved superior to introduced varieties, or to be valuable breeding material in combination with introduced varieties.” Further, this preservation of indigenous genetic diversity in regions of genetic diversity: “… does not infringe upon the development of the regions themselves”. The message here is clear: regions have the all the genetic diversity they need and therefore development is not needed. I reject both of these arguments: the first is only in a small part true; the second unethical (and as to ethics, what about the inevitability and danger of also conserving local and indigenous pests and diseases on-farm?).
Bennett (pp. 115-129) in the 1970 volume expands on the superiority of local populations. She believes in ‘local adaptation’, for example, as a result of disturbance associated with cultivation and contact with weed and wild relatives: “countless locally adapted races of cultivated species emerged.” And that: “local adaptive differentiation in primitive cultivars is general…”. We are therefore being told that local varieties, through local adaptation, are as good as it gets, and that introductions are not necessary. Bennett then makes the outlandish claim (p116): “Adaptation conferred by simple character differences has not been established by experiment. It is unlikely that adaptive traits can be attributed to single or even small number of genes.”
The package we are now presented with is that local varieties are best for farmers and that plants breeders can contribute little of value, as all adaptive changes are polygenic and beyond plants breeding. This massively ignores the work of pathologists (and Borlaug) in resistance breeding for stem and leaf rust in wheat, where ‘locally adapted races’ of wheat were so lacking in ‘local adaptation’ that all recent resistance genes have been sourced from wild relatives. It also conflicts with what we know of the non-optimality of local races (Gould, Namkoong) and the fact that introduced crops somehow do better than local ones (70% of crop production in Latin America and Africa is from crops introduced from other continents). It is also claimed in Frankel and Bennett (p. 12 – probably written by Bennett – Frankel knew better) that forest and pasture species: “may also perform substantially better than introduced material.” (p.12). As a generic belief, this is certainly wrong. Introduced forestry and pasture species regularly perform substantially better that native species, for example, Caribbean pine and eucalyptus all over the tropics and African pasture grasses in Latin America and Latin American pasture legumes elsewhere in the Tropics. And what about Parthenium? Is that locally adapted to wherever it grows?
But the message of Bennett’s questionable analysis of local adaptation is still with us. Local varieties of indigenous crops are thought to be best adapted and formal plant breeding and crop introduction unnecessary. This has been the belief and driving force of many on-farm projects, especially those run by NGOs that believe in farmer seed-saving as a mechanism to allow local adaptation (this ignores a fairly rapid varietal turnover on farm, a turnover that negates any possibility of local adaptation). It is this form of “yet more on-farm conservation” that I do not wish to see.
A further claim is that on-farm conservation is expected to allow crop adaptation to climate change – this could be like watching plasticine dry: far better to move crops and varieties around in regional on-farm trials. We already have millions of stored samples to deploy.
In contrast to the questionable premises of closed-system on-farm conservation there is the effectiveness of ex-situ conservation, feeding samples directly both to farmers elsewhere and to breeders. The dubious claims of progressive local adaptation and progressive change for characters of value can be tested by periodic collections on-farm and then ex-situ evaluation (and conservation) rather than by on-farm conservation projects. Frankel himself (in Frankel and Bennett, p. 476) commented on the questionable arguments of Bennett on the prospects of on-farm conservation: “But how is one to know? The outcome will depend on the dynamics of complex and uncontrollable components, resulting in changes that may well lead to massive erosion.” Frankel goes on to recommend conservation in dispersed experimental stations under scientific control. But these caveats have been ignored.
Also ignored are the multiple recommendations for a research agenda to develop a knowledge base for on-farm conservation. It is not possible to have meaningful projects without far more prior research. highlighted key areas for research. Recently Mercer and Perales (2010, in the context of climate change) gave a longer list of research needs. Why is this research not a priority? And where is the meta-evaluation of completed on-farm conservation projects?
Long ago I visited a community farming the Kalahari Sands of western Zimbabwe. On asking what farmers needed, a very old lady farmer said: “Crops and varieties from other countries”. This old lady, farming right on the edge, knew something Erna Bennett did not.
References
Brown A.H.D. (1999) The genetic structure of crop landraces and the challenge to conserve them in situ on farms. In S.B. Brush (ed.) Genes in the Field, International Plant Genetic Resources Institute, International Development Research Center, and Lewis Publishers, Rome, Italy, Ottawa, Canada and Boca Raton, FL, USA. pp. 29–48.
Mercer, K. L. and Perales, H. R. (2010) Evolutionary response of landraces to climate change in centers of crop diversity. Evolutionary Applications 3, 480-493.