Feeding the Future We Must Mine the Biodiversity in Seed Banks to Help to Overcome Food Shortages, Urge Susan Mccouch and Colleagues

COMMENT SPACE An elegy for the ART Exhibition revels GENES US Supreme Court OBITUARY Heinrich Rohrer, disappearing dark, in the power of patent rulings set a higher pioneer of scanning tunnelling banished by science p.26 unconstrained thought p.28 bar for innovation p.29 microscopy, remembered p.30 CARLOS ORTEGA/EPA/CORBIS CARLOS

The International Center for Tropical Agriculture in Colombia holds 65,000 crop samples from 141 countries. Feeding the future We must mine the biodiversity in seed banks to help to overcome food shortages, urge Susan McCouch and colleagues.

umanity depends on fewer than a landraces), as well as the wild relatives of crop food exporter. By 1997, the world economy dozen of the approximately 300,000 species and modern varieties no longer in use. had accrued annual benefits of approximately species of flowering plants for 80% These are stored in more than 1,700 gene $115 billion from the use of crop wild rela- Hof its caloric intake. And we capitalize on banks worldwide. Maintaining the 11 inter- tives2 as sources of environmental resilience only a fraction of the genetic diversity that national gene-bank collections alone costs and resistance to pests and diseases. resides within each of these species. This is about US$18 million a year. The time is ripe for an effort to harness the not enough to support our food system in The biodiversity stored in gene banks fuels full power of biodiversity to feed the world. the future. Food availability must double in advances in plant breeding, generates billions Plant scientists must efficiently and system- the next 25 years to keep pace with popula- of dollars in profits, and saves many lives. For atically domesticate new crops and increase tion and income growth around the world. example, crossbreeding a single wild species the productivity and sustainability of current Already, food-production systems are pre- of rice, Oryza nivara, which was found after crop-production systems. carious in the face of intensifying demand, screening more than 6,000 seed-bank acces- Why does plant breeding need a boost? climate change, soil degradation and water sions, has provided protection against grassy Because new, high-yielding seeds that and land shortages. stunt virus disease in almost all tropical rice are adapted for future conditions are a Farmers have saved the seeds of hundreds varieties in Asia for the past 36 years1. Dur- cornerstone of sustainable, intensified food of crop species and hundreds of thousands of ing the green revolution, high-yielding rice production3. Since the mid-1990s, progress ‘primitive’ varieties (local domesticates called and wheat varieties turned India into a net in conventional plant breeding has

slowed, despite the phenomenal yield samples in the world’s gene banks that are and fuelling plant improvement for years to gains of the past. Part of the reason is that available under the terms and conditions come. This requires an unprecedented effort only the tip of the biodiversity iceberg has of the ITPGRFA — perhaps up to 2 million. in data management and sharing. Today, seed been explored and used4. This ‘fingerprint’ for each plant will serve as data are typically recorded and managed by Crop wild relatives, landrace varieties the basis for assessing genetic relationships, different people, such as gene-bank curators, and previously undomesticated wild species and will make it possible to systematically agronomists and breeders, often in different represent sources of new variation for agri- select subsets of material for in-depth inves- institutions and in different database systems. culture. Such plants have survived repeated tigations. Sequencing costs are plummeting, But it is doable. The Global Biodiversity and extreme environmental challenges, yet making such an effort feasible. Information Facility (GBIF) — an online their resilience and adaptive capacity remain Sequence data provide a genomic ‘parts network that facilitates open access to “infor- largely untapped and poorly understood. list’ that can help to decipher mechanisms mation about the occurrence of organisms” A wealth of genetic information has been that enable plants to adapt to myriad envi- — provides a good example of such an infra- left behind throughout the history of plant ronments, and can structure and has changed how biodiversity domestication and scientific crop improve- “The research guide our remodelling is studied. But the GBIF does not currently ment. It must now be deployed. community of cropping systems handle the complex genomic data necessary Plant breeders often worry that using must pay for the future. Link- for our efforts. wild species or landrace varieties is too risky, attention to the ing sequence data Most importantly, results from genomics scientifically and economically. It took development with conventional and agronomic research must be connected 20 years and 34,000 attempts to cross a of locally ‘passport informa- to the communities that are creating new domesticated rice variety with a distantly adapted tion’ about collection varieties of crops. An international network related, highly salt-tolerant wild relative from varieties.” locality and original of scientists in both the public and the India before fertile offspring were obtained5. environment should private sectors must work together to It will now take at least 4–5 years of breed- call attention to the genetic potential of provide seeds and plants to farmers and ing to eliminate unwanted wild characters to many hidden crop resources. commercial plant breeders for further generate a new high-yielding, salt-tolerant Second, we must analyse the phenotypes of crossing and testing in different environ- rice variety (see go.nature.com/knztl5). That gene-bank accessions to evaluate their traits ments. The research community must is too long for most plant-breeding pro- and overall performance. This is the most pay specific attention to the development grammes, especially in the private sector. intellectually challenging, complex, costly of locally adapted varieties that meet the Insufficient genetic and phenotypic infor- and time-consuming stage. We cannot hope needs of the world’s poorest farmers. mation about most of the holdings in gene to evaluate all gene-bank accessions in all How much would such a systematic, banks makes plant breeders even more relevant environments, even with the advent concerted, collaborative global effort to reluctant. Politics has also created obstacles. of high-throughput phenotyping technolo- feed the future cost? We estimate around The Convention on Biological Diversity (see gies. Using sequence data in combination $200 million annually. This seems like great go.nature.com/njehon) is an international with phenotypic, geographical and ecological value, given that as a society we have spent treaty that, although vital for consolidat- information will enable researchers to target $3 billion on sequencing the human genome, ing efforts to conserve the diversity of life field experiments strategically and to develop $9 billion on constructing CERN’s Large on Earth, has created significant barriers to models that can predict plant performance. Hadron Collider near Geneva in Switzer- the sharing of genetic material, including of This will make plant breeding faster, more land (plus about $1 billion a year in running domesticated plants and their wild relatives4. efficient and cheaper. costs) and can spend up to $180 million on Happily, things are changing. The Interna- Assessing the breeding potential of unfa- a single fighter jet. tional Treaty on Plant Genetic Resources miliar plant materials typically requires them After all, as the ecologist Charles Godfray for Food and Agriculture6 (ITPGRFA), to be crossed with modern, ‘elite’ varieties. put it: “If we fail on food, we fail on every- negotiated in 2004, now governs access to Their offspring are then evaluated in envi- thing”. ■ crop diversity. It mandates that a portion ronments of interest to farmers and breeders, of any monetary benefits derived from over several years. Often, the genetics of high- Susan McCouch is professor of plant the commercialization of products from performing offspring can be traced back to breeding and genetics and of plant biology gene-bank materials is put into a fund that DNA inherited from wild or landrace donors at Cornell University, Ithaca, New York. supports conservation and sustainable use that are agriculturally less productive. For e-mail: [email protected] of crop genetic resources. example, the wild tomato species Solanum On behalf of attendees and organizers of the On the technical front, we are now able pennellii was used to double commercial Crop Wild Relative Genomics meeting held use a plant’s genetic make-up to predict its tomato yields under a wide range of growing in Asilomar, California, in December 2012. agronomic potential and traits. Plant breed- conditions7, and the wild rice species Oryza See go.nature.com/nrpoe3 for full author list. ers commonly use genetic markers to identify rufipogon increased yields of elite varieties of individual plants carrying specific genes for rice by more than 25% (ref. 8). Thus, useful 1. Plucknett, D. L. et al. Gene Banks and the World’s Food (Princeton Univ. Press, 1987). disease and pest resistance or stress tolerance, genetic traits are moved across the breed- 2. Pimentel, D. et al. BioScience 47, 747–757 without ever exposing the plants to the rel- ing barrier, expanding the genetic diversity (1997). evant agents. Breeders can use genome-wide of domesticated plants and opening up new 3. Godfray, H. C. J. et al. Science 327, 812–818 (2010). approaches to eliminate 70–80% of individu- opportunities for environmental resilience 4. Tanksley, S. D. & McCouch, S. R. Science 277, 4 als in any generation without having to invest and future gains in quality and yield . 1063–1066 (1997). in laborious multi-environment field testing. A third key step is to create an internation- 5. Jena, K. K. et al. Rice Genet. Newsl. 11, 78–79 (1994). ally accessible informatics infrastructure to 6. Food and Agriculture Organization. International THREE STEPS catalogue the diversity in the world’s seed Treaty on Plant Genetic Resources for Food and How should we begin to mine biodiversity collections. This would link seeds and genetic Agriculture (FAO, 2009); available at http:// for food security? A logical first step is to stocks directly to passport, genomic and go.nature.com/ikflqh. 9 7. Gur, A. & Zamir, D. PLoS Biol. 2, e245 (2004). obtain a sample of sequence information phenotypic information , thereby engag- 8. Imai, I. et al. Mol. Breed. 32, 101–120 (2013). from the genomes of all non-duplicate plant ing the creativity of geneticists and breeders 9. Zamir, D. PLoS Biol. 11, e1001595 (2013).