What a Seed Bank Holds

A Narrowing World

There is a photograph taken in 1893 at a seed trial in Geneva, New York, that shows 408 distinct varieties of tomato arranged on display tables, each one different in color, shape, texture, and ripening behavior. Some were bred for humid summers in the mid-Atlantic. Some for the short cool seasons of Vermont. Some were the product of generations of selection by farm families who passed their seeds down the way other families pass down furniture or recipes, as irreplaceable pieces of a particular way of living in a particular place.

Most of those varieties are gone now. Not just from that field trial, but from the earth entirely.

The Food and Agriculture Organization of the United Nations estimates that roughly 75 percent of crop genetic diversity was lost during the 20th century, as farmers around the world replaced locally adapted, open-pollinated varieties with the high-yielding, genetically uniform crops promoted by industrial agriculture.[1] The number of crop varieties actively grown on farms, not merely stored in freezers, fell by more than 90 percent over that same period.[2] Today, just nine plant species account for 66 percent of global crop production, and rice, wheat, and maize alone provide more than 50 percent of the world’s plant-derived calories.[3] A food system that once drew on thousands of varieties of dozens of species now depends, functionally, on a handful.

This is not a historical curiosity. It is a structural vulnerability. Each lost variety represents a set of traits, drought tolerance, disease resistance, cold hardiness, particular nutritional profiles, particular flavor compounds, that cannot be recreated once the seed is gone. Unlike almost any other form of knowledge, genetic diversity cannot be reverse-engineered. When a seed line dies out, it dies out permanently. The loss of a crop variety, as the Crop Trust notes, “is as irreversible as the extinction of a dinosaur, animal, or any form of life.”[4]

How Industrial Agriculture Selected Against Diversity

The mechanics of this loss are not mysterious. They follow directly from the economics of industrial food production, and from a set of breeding priorities that had nothing to do with flavor, nutritional density, regional adaptation, or the long-term resilience of the food system.

Begin with the tomato. The commercial tomato market in the United States is worth billions of dollars annually. The average commercial tomato travels roughly 1,500 miles from field to consumer, spending days in refrigerated trucks, weeks in distribution centers, and additional time on store shelves before anyone eats it. This supply chain imposes a very specific set of requirements on the tomato itself: it must be uniform in size and shape for mechanical harvesting, firm enough to survive transport without bruising, with a thick enough skin to resist damage, and capable of ripening slowly and predictably after being picked green. Flavor, the concentration of sugars, acids, and volatile aromatic compounds that makes a tomato taste like a tomato, is largely irrelevant to these requirements. In many cases, the traits associated with flavor and nutritional density are actively selected against, because they correlate with softer texture and faster deterioration.[5]

The same logic plays out across the produce aisle. Commercial apple varieties are selected for appearance, shelf life, and shipping durability. Commercial strawberries are bred for size and firmness. Commercial corn is bred for yield and uniformity, not for the complex carbohydrate profiles or flavor characteristics of the hundreds of indigenous varieties it largely replaced. Over seventy years of industrial plant breeding, the accumulated result of these selection pressures was a radical narrowing of the genetic base of our food supply.

Concentration in the seed industry accelerated the process. Today, four agrochemical companies, Bayer (which acquired Monsanto), Corteva, Syngenta, and BASF, control approximately 60 percent of the global seed market.[6] Before the major mergers of the early 2000s, there were hundreds of independent regional seed companies, many of which maintained and sold varieties bred specifically for local growing conditions. As these companies were acquired, their catalogs were often consolidated and narrowed, eliminating varieties that were commercially marginal but ecologically irreplaceable. Where farmers once had access to seed specifically adapted to their climate, soil, and market, they increasingly found themselves choosing from a short list of globally optimized varieties designed for commodity production at scale.

The irony is that the very diversity that industrial breeding discarded is what climate resilience requires. A food system that depends on a narrow genetic base, however high-yielding that base may be, is inherently fragile. A new pathogen, a shift in rainfall patterns, an unexpected pest can devastate a monoculture of genetically similar plants in ways that a diverse polyculture can survive. The potato famine that killed more than a million Irish people in the 1840s was, at its root, a biodiversity failure: an entire nation’s food supply built around a single clonal variety, with no genetic reserve to fall back on when blight arrived.

“The loss of a crop variety is as irreversible as the extinction of a dinosaur, animal, or any form of life.” — Crop Trust, Svalbard Global Seed Vault

The Global Response: Seed Vaults and Genebanks

The international scientific community has been aware of the biodiversity crisis for decades. The response, at the largest scale, has been the construction of a global network of genebanks: institutions that collect, catalog, and store seeds under controlled conditions, preserving genetic material that might otherwise vanish entirely.

Svalbard: The World’s Doomsday Garden Shed

The Svalbard Global Seed Vault, located 1,300 kilometers beyond the Arctic Circle inside a sandstone mountain on the Norwegian island of Spitsbergen, is the most famous of these institutions. Opened in 2008 and managed jointly by the Norwegian government, the Nordic Genetic Resource Center, and the Crop Trust, it functions as a backup for the world’s entire genebank network: a facility where any genebank on earth can store duplicate copies of its collection, retrievable if the original is destroyed by war, natural disaster, funding collapse, or simple equipment failure.[7]

As of late 2025, the Vault holds more than 1.3 million seed samples originating from nearly every country in the world, contributed by 130 genebanks across more than 90 countries.[8] The facility has the theoretical capacity to store 4.5 million varieties, roughly 2.5 billion individual seeds, and remains, by most measures, the most secure seed repository ever built. It has already been drawn upon once in anger: in 2015, the International Center for Agricultural Research in the Dry Areas (ICARDA), forced to relocate from Syria due to the civil war, withdrew a portion of its collection from Svalbard to reestablish its genebank in Lebanon and Morocco, the first withdrawal in the facility’s history, and a stark demonstration of exactly why it exists.[9]

The network that feeds Svalbard is itself substantial. As of 2024, an estimated 1,750 genebanks worldwide hold collections of approximately 7.4 million varieties of plants, an enormous archive of genetic material.[10] The Vavilov Institute of Plant Industry in Russia, the oldest genebank in the world, holds more than 320,000 accessions. The U.S. National Plant Germplasm System maintains hundreds of thousands more. The CGIAR network of international agricultural research centers collectively stewards some of the most critical collections on earth, including the rice diversity held at the International Rice Research Institute in the Philippines and the wheat and maize collections at CIMMYT in Mexico.

The Limitation No Freezer Can Solve

And yet, the scientists who run these institutions are among the first to acknowledge their limits. A seed in cold storage is preserved, but it is not alive in the fullest sense. It is not adapting. It is not being selected by the conditions of a real growing season. It is not co-evolving with the soil microbiome of a specific place, responding to the particular pest pressure of a particular summer, producing offspring that carry the slight genetic adjustments that come from actually being grown.

This distinction matters enormously. A seed variety stored at minus 18 degrees Celsius in a Norwegian mountain is safer from physical loss than one growing in a garden in Georgia. But the garden variety is doing something the frozen seed cannot: it is living. Plants evolve in response to where they are grown. A tomato variety grown in South Florida for five generations will gradually accumulate adaptations to South Florida’s heat, humidity, and pest pressure that make it more productive there than a genetically identical variety that has spent those same years in a freezer. This process is called in situ conservation, preservation through actual cultivation, and it is, ultimately, irreplaceable.

The Crop Trust itself, which manages the Svalbard network, explicitly acknowledges this. Its conservation model calls for genebanks to regularly “regenerate” their collections, to grow seeds out, harvest new ones, and return them to storage, specifically to maintain viability and genetic integrity. Without that cycle of actual growing, even a perfectly maintained genebank gradually loses the living relationship between seed and soil that makes genetic diversity meaningful.

The American Seed Sovereignty Movement

While global institutions have worked to preserve seeds in cold storage, a parallel movement has been building in American gardens and on American farms for fifty years, one that takes a fundamentally different approach to the problem. Rather than extracting seeds from the communities and landscapes that produced them and housing them in centralized facilities, the seed sovereignty movement keeps seeds in circulation: in active gardens, in community exchanges, in the hands of the people who grow food for their families and their neighbors.

Seed Savers Exchange: The Anchor Institution

The story of organized seed saving in America begins in 1975, with two seeds. Diane Ott Whealy’s grandfather, Baptist John Ott, entrusted to her the seeds of two varieties he had kept for generations: Grandpa Ott’s morning glory and the German Pink tomato, both brought from Bavaria by his parents when they immigrated to Iowa in 1884. Recognizing that these seeds, and the thousands of similar heirloom varieties held quietly by gardeners and farm families across the country, were at risk of vanishing as industrial agriculture contracted the commercial seed supply, Diane and her husband Kent formed a network for preserving and sharing them.[11]

That network became Seed Savers Exchange (SSE), now headquartered on 890 acres in Decorah, Iowa, and operating as the largest nongovernmental seed bank in the United States. SSE stewards a collection of more than 20,000 heirloom and open-pollinated varieties, vegetables, herbs, and fruits, at its Heritage Farm, and maintains a backup deposit at the Svalbard Global Seed Vault. In 2025, SSE celebrated its 50th anniversary.[12] Its commercial catalog offers over 600 varieties for sale; its member network provides access to thousands more through the annual yearbook seed listing, through which member gardeners offer seeds they have grown and saved. The exchange has distributed more than one million seed samples to date.

What SSE represents is not just a seed collection. It is a philosophy of preservation built on what the organization calls “participatory preservation”, the understanding that seeds are most fully protected when they are being actively grown by people who care about them. A seed in SSE’s gene bank is safe. A seed being grown in a member’s garden in Alabama, adapting to Alabama summers, selected each year by a gardener who knows that land is safer still.

Native Seeds/SEARCH: Preserving Indigenous Arid-Land Crops

Founded in 1983 in Tucson, Arizona, Native Seeds/SEARCH (NS/S) emerged from a specific crisis: researchers working with the Tohono O’odham Nation on food security discovered that traditional crops, foods that Indigenous communities had grown and depended on for centuries, specifically adapted to the brutal heat and limited rainfall of the Sonoran Desert, were rapidly disappearing from active cultivation. NS/S was founded to find, document, and preserve them.[13]

Today, NS/S stewards approximately 1,900 accessions of traditional crops in a climate-controlled seed bank in Tucson, with 70 percent of those varieties originating directly from Indigenous communities of the American Southwest and northwest Mexico.[14] The collection includes drought-adapted corn varieties, native beans, desert-adapted squash, chia, amaranth, and hundreds of other species representing millennia of agricultural knowledge developed in one of the most demanding growing environments on the continent. NS/S distributes seeds for free to Indigenous growers with cultural and historical connections to the collection, and sells through an online catalog to support conservation work.

The organization’s central insight, that seeds belong to the communities that developed them, and that those communities are the most appropriate stewards of them, has influenced seed conservation practice well beyond the Southwest.

Seed Companies Doing It Right

Beyond the nonprofit world, a growing constellation of independent seed companies has built businesses explicitly around the values of open pollination, regional adaptation, and genetic diversity. These are not companies that happen to sell heirloom seeds as a niche product. They are companies that have structured their entire operations around the principle that seeds should remain in the commons, adapted to the places where they are grown, and accessible to the farmers and gardeners who need them.
A note on open pollination and why it matters: open-pollinated varieties are those that reproduce true from seed when properly isolated. A gardener who saves seeds from an open-pollinated tomato can plant those seeds next year and get the same tomato. Hybrid seeds, designated F1 on most commercial packets, do not breed true; they must be purchased new each season, creating a dependency on the commercial seed supply that open-pollinated varieties do not. As Adaptive Seeds puts it, proprietary hybrids are “the single biggest contributor to the erosion of agricultural biodiversity regionally and worldwide.”[15] Open pollination is not nostalgia. It is food sovereignty.