Genomics and genebanks

The onward rush of genomics technology continues apace. Today, the perceived constraints of its widespread adoption in plant breeding programmes – such as a paucity of informative markers, or an economically sustainable means of extracting DNA – seem to be rather quaint notions. Published just before the turn of the millennium, Young's (Reference Young1999) ‘cautiously optimistic vision for marker-assisted selection’ now feels dated and overcautious, and the idea of introducing a major component of early-generation genotypic selection (‘plot selection or plate detection?’) is not quite as outlandish as it may have seemed to Koebner and Summers (Reference Koebner and Summers2003). The question concerning how these technologies will alter conventional paradigms of pre-breeding and genebanking is therefore of current interest. The key, as exemplified by Susan McCouch in her exposition of the state of the art in rice (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Susan_McCouch.pdf), is the reliability of using genotype as a surrogate of phenotype and, specifically, of performance with respect to the quantitative traits of most relevance for crop improvement and adaptation. The small size and high level of documentation of the rice genome offer the opportunity to think ‘big’ in this case, to consider both the re-sequencing of a collection of 125 cultivars and near wild relatives, generating an ocean of data describing nucleotide variation at 16 million sites, and the single nucleotide polymorphism (SNP) genotyping of around 2000 accessions at just (!) 44,000 of these sites to allow genome-wide association studies. From a pre-breeding perspective, the hoped-for outcomes of this type of study will be the recognition of unexpected donors for specific traits and the directing of allele mining among as yet underutilized genetic resources.

The sudden availability of oceans of data in itself is forcing a change in the mentality of the genebank curator. The day when full genome sequence will be to hand for most, if not all, of our crop species, and thereafter, pace rice, for multiple accessions of each, is arguably not so far off. The question addressed by Rob van Treuren was how all these sequence data will and/or should impact upon the management and operation of germplasm collections (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Robbert_van_Treuren.pdf). For the moment at least (although surely this will develop over time), some genebanks are sufficiently techno-savvy to deal with the issue, while others are not; but even with respect to the former, it is unclear whether the mandate of a genebank extends to deriving (as opposed to merely curating) sequence data, leaving the user community to provide these data. The present authors at any rate would suggest that as genotype is already an integral component of the passport of a germplasm entry (and will only become more so), then germplasm conservators will have little option but to involve themselves in the acquisition of these data, where they can. The implications on future genebank practice are wide ranging, but no more so than the effect which ‘omics’ technologies have had on the conduct of biological research in general. The challenge remains to link variation at the levels of gene sequence and of metabolic pathways with useful plant genetic resources for the purpose of accelerating cultivar development.

Food security

In a world where already an estimated one billion are chronically hungry (Shapouri et al., Reference Shapouri, Rosen, Meade and Gale2009), securing the supply of food is an issue that demands to be tackled now (Conway, Reference Conway2012). From a genebank perspective, the challenge is to identify accessions that meet the needs of plant breeders. As pointed out by Ruraidh Sackville-Hamilton (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Ruaraidh_Sackville_Hamilton.pdf), it is an article of faith that, hidden among all the millions of accessions curated in local, national and international collections, are carriers of genes which could improve yield potential, adapt to a changed climate, improve nutritional quality, resist damaging disease(s) and pest(s) and even make a positive contribution to agricultural sustainability. But where? And how to find them? The history of plant breeding suggests that a level of optimism would not be misplaced. A recent cost–benefit study of the impact of breeding conducted by the International Rice Research Institute (IRRI) in just three countries (the Philippines, Indonesia and Vietnam) has estimated that every dollar invested has returned over US$22, a rate of return that would be the envy of any industrialist. Several thousand current and recent rice cultivars include IRRI material in their pedigree. The awkward truth is, however, that the vast bulk of the accessions curated in the IRRI genebank has not contributed at all to this economic gain. A rationalist's view might therefore be that there is little to be lost in discarding rarely (or never) used accessions, but this argument is neatly scotched by the example of the landrace FR13A, collected in the 1950s, identified as submergence-tolerant in the late 1970s and exploited for the release of cultivars in the late 2000s. The pace of progress will have to be accelerated, however, to keep up with population growth and climate change, and the main bottleneck remains phenotypic characterization. It is hardly feasible to mass phenotype whole collections, even for one trait, let alone for many. The most productive way forward is therefore seen to lie in achieving a quantum leap in predictive genotyping.

Two further examples of the impact of genomic data on characterizing major food crops were presented by Sara Hearne in maize (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Sarah_Hearne.pdf) and Audrey Didier in wheat (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Audrey_Didier.pdf). The Mexican government-funded programme Seeds of Discovery is targeting the 120k wheat and 40k maize accessions in the CIMMYT genebank for genotyping by sequencing. The intention is to shift useful traits from non-adapted germplasm into material which a breeder can adopt without incurring the risk of too much genetic drag. The key to this programme is to test for genotype/phenotype associations on a massive scale (e.g. making test crosses between a double haploid inducer line and 4000 extensively genotyped accessions also phenotyped for a wide range of abiotic, biotic and quality traits). The parallel wheat programme is still at the stage of acquiring mass phenotypic data from large field trials. Meanwhile, the French public/private partnership project Breedwheat is seeking to introduce novel diversity into the elite gene pool by searching for genotype/phenotype associations directed at abiotic and biotic stress tolerance. A continuing research activity focuses on identifying patterns in the distribution of genetic diversity, with a view to enabling the selection of accessions to meet plant breeding needs and to designing methods that optimize a pipeline leading from genebank accessions to ‘omics-led’ ideotype breeding.

Nutrition and health

The Green Revolution's achievement was to double wheat and rice production between 1966 and 1990, at a time when the population of densely populated low-income countries was growing by 80% (Khush, Reference Khush1999). The prediction is that by 2000, food prices would have been about 19% higher than they actually were without the interventions led by the international agricultural research centres, and that per capita calorie consumption would have dropped by about 5%, resulting in a 2% or greater increase in the number of malnourished children (Evenson and Gollin, Reference Evenson and Gollin2003). The absolute focus was, rightly therefore, to increase the supply of calories. Plants can of course provide much more than calories; somewhere between 20 and 50% of the 1.2 billion people of India also rely heavily on plant products for their intake of protein, vitamins and essential minerals. Meanwhile, there is a growing realization that current levels of meat consumption in more developed economies are becoming untenable, given the calorific inefficiency of its production, along with the potential environmental hazards associated with its intensified production. We feed livestock nearly as much grain as we feed ourselves, and some 33% of cropland is currently dedicated to the production of animal feed. The challenge for the plant genetic resources community is therefore to exploit genetic variation relevant for enhancing plant-based food quality and functionality. As Hilde Nybom reminded us, plant breeding ‘is not all about starch’ (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Hilde_Nybom.pdf); for example, the content of specific phenolics in the apple peel varies notably among cultivars, and a whole spectrum of fruits and berries offers a largely underutilized source of antioxidant molecules. For the breeder, unfortunately, the genetic basis of the variation (both intra- and interspecific) in antioxidant content is complex and is also heavily influenced by the growing environment; a similar story applies to the content of fatty acids, flavonoids and resveratrol in groundnut. The genetic basis of the content of the many (76) components contributing to aroma in bread wheat, as shown by Jihad Orabi (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Jihad_Orabi.pdf) is particularly challenging: 267 putative quantitative trait loci were identified! Meanwhile, in grape, the use of high-density marker technology has identified a few candidate genes responsible for the flavour of the moscatel type (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Valeria_Terzi.pdf). The complexity of end-use quality traits only emphasizes how crop improvement with respect to enhancing human health and nutrition requires a strong interdisciplinary approach, taking into account biochemistry, medicine and nutrition – not to mention sociology.

Contribution of crop wild relatives

The potential of crop wild relatives (CWRs) as a source of genes for crop improvement has long been recognized, and has indeed been realized to an extent in some crop species (Hajjar and Hodgkin, Reference Hajjar and Hodgkin2007). An example that has surfaced in the popular press of late relates to the resynthesized hexaploid wheats championed by CIMMYT (Warburton et al., Reference Warburton, Crossa, Franco, Kazi, Trethowan, Rajaram, Pfeiffer, Zhang, Dreisigacker and van Ginkel2006), over-hyped perhaps by the BBC as ‘superwheat’ (http://www.bbc.co.uk/news/uk-22498274). Hannes Dempewolf from the Global Crop Diversity Trust rather colourfully illustrated the dilemma faced in making interspecific crosses, by pointing out that making a cross between a house cat and a wild cat may not produce a docile kitten, but rather some creature which one probably would not want to have lying on one's sofa (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Hannes_Dempewolf.pdf); he might equally have described Georgii Karpechenko's hybrid between radish and cabbage, which unfortunately combined the cabbage's root with the radish's flower head, rather than the desired reverse!

The conservation of CWRs is of major concern; while ex situ approaches are being pursued by the Global Crop Diversity Trust and the FAO (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Jose_M_Iriondo.pdf), in situ ones were argued by Nigel Maxted (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Nigel_Maxted.pdf) to be even more important than they are for the crop species themselves. While substantial investment has been made for the ex situ conservation of crop cultivars, a major concern was expressed for the survival of crop landraces, for which a Europe-wide scheme for on-farm conservation was described by Valeria Negri (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Valeria_Negri.pdf).

Traits relevant for addressing climate change

Addressing the problem of climate change is complicated by the realization that we are battling an uncertain storm. The evidence that it is occurring as a result of anthropogenic emissions of greenhouse gases (GHGs) is convincing enough to all, but the swivel-eyed sceptic (Stern, Reference Stern2007), although exactly how this will play out in terms of local and global temperature rise, climatic unpredictability (in particular rainfall) and sea level rise, remains uncertain. The yield potential of maize, soybean and cotton has been estimated to be compromised by 6–10% per every 1°C over 30°C. Climate modelling, according to Andy Jarvis (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Andy_Jarvis.pdf), suggests that 30% of the world will experience altered climatic conditions by 2050 (although not all of these changes will necessarily have a negative effect on crop productivity). The message is clear – because the development of new cultivars is typically slow (for example, the time elapsed between the release of a winter wheat cultivar and the making of the original progenitor cross is seldom less than a decade), the moment to be searching for adaptive variation in plant genetic resources is now! The key question is how to select for adaptation to an environment that does not exist at present. The potential of focused identification of germplasm strategy was illustrated by Hamid Khazaei (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Hamid_Khazaei.pdf), who described the search for drought tolerance in faba bean by screening materials originating from an area that currently experiences a climate similar to that predicted to occur in the major production zones. A more direct approach for identifying adaptation in Nordic barley to the combination of higher temperature, higher carbon dioxide concentration and higher ozone concentration was taken by Catherine Ingvordsen (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Cathrine_Heinz_Ingvordsen.pdf); this experiment of course required access to a rather sophisticated phytotron facility, which has also been exploited for similar experiments involving Brassica spp. (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Rikke_Bagger_Jorgensen.pdf). In all cases, genetic variation for adaptation was exposed and better-performing accessions identified.

Reducing the footprint of agriculture

Although an awareness of agricultural sustainability dates back to some of the most ancient writings in both Europe and Asia, the modern pre-occupation with it is focused on the need to reduce the effect of farming on the environment by developing technologies that are both practical and affordable, seeking to simultaneously improve productivity and ameliorate the environment (Pretty, Reference Pretty2008). The enhancement of yield achieved in the Green Revolution may have been associated with an increased level of GHG emission associated with higher fertilizer production and application, but, overall, its net effect has been calculated to have reduced CO₂ emission by some 161 Gt over the period 1961–2005 (Burney et al., Reference Burney, Davis and Lobell2010), implying that gains in crop productivity can make a positive contribution to reducing GHG emission. A major role of genetic resources will be to provide germplasm less dependent on external inputs to maintain current levels of productivity. As explained by Matthew Hegarty (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Matthew_Hegarty.pdf), grazing cattle typically excrete 55–95% of the nitrogen and 20–70% of the phosphorus that they ingest, while 2–12% of gross energy intake is lost as methane. Much of this waste ends up as environmental pollution. Breeding for higher digestibility in rye grass is seen as a means of mitigating some of these losses, and a key to this property is the content of water-soluble carbohydrate present in the forage. The application of high-density genotyping is yielding possible candidate genes for this trait. Ari Rajala (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Ari_Rajala.pdf) described a historical study of the effect of breeding on nitrogen use efficiency in barley. While breeding has obviously improved yield, modern cultivars tended to out-perform older landraces with respect to nitrogen use efficiency. Nevertheless, certain landraces did show particularly unexpectedly high levels of nitrogen utilization or uptake efficiency, while at the same time some progress was made towards identifying SNP markers associated with these traits. A priority will be to provide evidence to policy makers that plant genetic resources (and their enhancement through plant breeding) are the key for ‘green growth’ in a bio-based economy under an uncertain climate.

Impact of intellectual property on genebanks and public/private partnerships

The final session of the meeting addressed the vexed issue of intellectual property in the context of plant genetic resources. In the far-off good old days, germplasm could be exchanged without restrictions or obligations, and the idea of a ‘material transfer agreement’ would have seemed as incomprehensible to plant genetic resources practitioners as was the creature that confronted Ellen Ripley in Alien. Following the 1992 Convention on Biological Diversity and the signing of the International Treaty on Plant Genetic Resources for Food and Agriculture in 2004, access has become more regulated. An impassioned argument from the viewpoint of the private breeding sector was made by Anke van den Hurk (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Anke_van_den_Hurk.pdf), who stressed that the net effect of the legislation has been to inhibit rather than to encourage the use of genetic resources. A similar exposition from the public viewpoint was put by Theo van Hintum (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Theo_van_Hintum.pdf). Despite this rather sombre scenario, the concept of public/private partnerships in pre-breeding projects was viewed quite positively. Ahmed Jahoor described one such programme focused on barley (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Ahmed_Jahoor.pdf), Odd-Arne Rognli the same for ryegrass (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Odd_Arne_Rognli.pdf) and Dag Røen for apple (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Dag_Roeen.pdf), all carried out in the framework of Nordic cooperation. As an appropriate full stop to the meeting, Øla Westengen reminded us of all of the activities of the huge natural fridge that is the Svalbard Global Seed Vault (http://www.nordgen.org/ngdoc/nordgen/Eucarpia/oral_presentations/Ola_Westengen.pdf).