In 1868, Charles Darwin published an erudite and influential book — not On the Origin of Species, which had been in print for nine years already, but rather The Variation of Animals and Plants Under Domestication. In this work, Darwin takes a long and careful look at the range of varieties of domesticated organisms. He explores both how they differ from their wild relatives and how these variations have been brought about by the activity of people. Although there is a lot about pigeons, as Darwin was a great pigeon fancier and used the species to carry out many breeding experiments, he doesn't neglect domestic crops. Chapters are devoted to cereals and culinary plants, fruits and flowers, and there is one ‘On bud-variation, and on certain anomalous modes of reproduction and variation’.

In this book, Darwin identified two truths that continue to occupy us today: that “cultivated races of plants often exhibit an abnormal character, as compared with natural species; for they have been modified not for their own benefit, but for that of man”, and that “sterility often supervenes when organic beings … are too closely interbred”. In other words: selective breeding is geared towards traits that satisfy the short-term desires of farmers and consumers, but may not be of long-term benefit. Look, for example, at the iceberg lettuce: bred for crispness and shelf life, it now has so little nutritional value that rabbit owners are warned not to feed it to their pets.

What Darwin didn't and couldn't know was that artificial selection cannot necessarily reverse a decline in useful characteristics. Some traits, once lost, cannot be brought back despite any amount of crossing of different domestic strains. One reason stems from the tendency of crop plants to be polyploid, the expansion of their genetic material often being closely associated with the earliest steps in the domestication process. Bin Han of the Chinese Academy of Sciences in Shanghai explores such processes in a News and Views article this month (16042), looking at a recent genetic study of peanuts (D. J. Bertioli et al. Nature Genet. 48, 438-446; 2016). Becoming polyploid by virtue of hybridization of two different species immediately creates a reproductive barrier between the new hybrid and the parental species from which it is derived. If the hybrid is fertile, this foundation event will remain an evolutionary bottleneck, cutting off future generations from the genetic diversity from which it emerged.

Faced with such domestication-induced genetic impoverishments, where can we turn to find lost traits that an ever-changing environment may have made highly advantageous? If domestication has resulted in a bottleneck that cuts off a crop from the genetic diversity of its wild parents, then the existing populations of those progenitors could be a good place to start looking.

And that is exactly what Vivianne Vleeshouwers and colleagues reported in Nature Plants last year (15034). They tranferred the gene for an elicitin receptor from the wild potato Solanum microdontum into a strain of domestic potato (Solanum tuberosum L.), thus conveying resistance to the oomycete Phytophthora infestans, the causative agent of the highly destructive late potato blight.

The wild relatives of our domestic crops, as well as heirloom varieties, are an important genetic resource for future crop breeding, and must be preserved. A News Feature in this issue (16048) discusses initiatives by organizations such as the International Rice Research Institute in the Philippines and the International Potato Centre in Peru that aim to maintain and exploit the genetic heritage of crops. Meanwhile in an Article in this issue (16022), Nora Castañeda-Álvarez and colleagues attempt to quantify the global conservation status of crop wild relatives, and thus set priorities for future collection expeditions to prevent too much of this priceless genetic resource being lost before its true value has been determined.

This opens the question of how best to maintain the reservoir of crop wild relatives and heirloom varieties. One answer is seed banking, as seeds of most plants are relatively easy to store for long periods by drying and freezing. In recent years, a global network of seed banks has been established by the collaboration of botanic gardens and research institutes. Small local banks are responsible for collecting and preserving varieties from their particular areas, and larger establishments provide facilities to store samples from across the globe. At the top of this pyramid is the Svalbard Global Seed Vault in Norway. A mere 600 miles from the North Pole, it is designed to act as a bank of last resort, capable of protecting its treasure house of seeds for a thousand years whatever disasters, natural or otherwise, might occur.

But not all crops can be stored as seeds. Fruits and other long-lived perennials exhibit self-incompatibility, and so their varieties can only be maintained genetically pure by vegetative propagations. Their survival requires that growing collections be maintained. A good example would be the UK's National Fruit Collection at Brogdale Farm in Kent. Their more than 2,000 apple varieties, 500 pears, 1,000 stone fruits and an assortment of nuts are held alongside detailed phenotypic data. Cryopreservation also has a role, although it is buds rather than seeds that are frozen and resurrected through grafting to recover lost specimens.

Living collections also preserve something less tangible than the plants themselves: an understanding of how to grow and cultivate these crops. Much of this knowledge is maintained within the cultural traditions of the peoples who rely on them, whether Kentish cherry growers or Bolivian quinoa farmers. The World Intellectual Property Organization defines this as ‘traditional ecological knowledge’, and UNESCO covers it within its Intangible Cultural Heritage programme. However, such endeavours often equate ‘safeguarding’ with ‘recording’; once a record exists, the experience in question is ‘safe’. But even the most thorough description of a dodo will not allow us to bring it back to life.

However important the work of seed banks or collections might be — and they are indeed vital — agriculture is more than just the plants that support it. If we are to have the resources of heritage varieties and crop wild relatives available to counter the problems of future food security, we need to maintain them in their native environments, tended by the peoples who have relied on them for millennia.