Why Crop Diversity Collapsed and Why That Matters

In the autumn of 1845, a fungal pathogen called Phytophthora infestans swept through the potato fields of County Mayo, Ireland, and turned the entire harvest to black slime within days. The crop that fed roughly a third of Ireland’s population was not merely damaged — it was annihilated. What followed was not only a humanitarian catastrophe that killed a million people and sent two million more fleeing across the Atlantic. It was the first modern demonstration of a principle that agricultural civilization has never quite accepted: monoculture is a debt instrument. You borrow stability from the future, and eventually the bill arrives.

The Irish were not growing many kinds of potato. They were growing one — the Lumper, a single variety that yielded enormous quantities of starchy calories on poor soil. It spread across Irish tenant farms not because it was the best potato in any biological sense, but because landlords and markets rewarded volume above all else. Every farmer growing the same plant made the entire system identically vulnerable to the same pathogen. When Phytophthora arrived, it found a continent-spanning monoculture and treated it accordingly.

We have been repeating this experiment at larger and larger scales ever since.

The Logic That Produced Uniformity

To understand why crop diversity collapsed, you have to understand the incentive structure that replaced it. For most of human agricultural history, farmers were their own seed stewards. They saved seed from their best plants, traded locally, selected for the specific conditions of their soil and microclimate, and maintained what plant geneticists now call “landraces” — populations of a crop with enormous internal genetic variation, adapted over generations to particular places. A single landrace of wheat in Turkey or a traditional corn variety in Oaxaca contained more genetic diversity than the entire commercial seed supply of a modern nation.

This system was locally optimal and globally robust. A blight that devastated one village’s variety might spare the next valley’s, because the two populations were genetically distinct. The system was also economically invisible — farmers captured no market value from their breeding work, no intellectual property accrued, no one aggregated the value of the diversity being maintained. It was a public good produced as a byproduct of subsistence farming.

Then the Green Revolution arrived. Beginning in the 1940s and accelerating through the 1960s, plant breeders at institutions funded by the Rockefeller and Ford Foundations developed semi-dwarf wheat and rice varieties with extraordinary yield potential — provided they received sufficient fertilizer, irrigation, and pesticide inputs. Norman Borlaug’s wheat varieties spread from Mexico to Pakistan to India, and grain yields doubled and tripled in places that had faced recurring famine. The humanitarian case for the Green Revolution is real and I will not dismiss it. Hundreds of millions of people did not starve who otherwise would have.

But the mechanism of that success was standardization. The new high-yield varieties worked best when every farmer in a region planted the same thing at the same time with the same inputs. Local variety and heterogeneity were not just economically irrelevant — they were agronomic obstacles. Extension services pushed monoculture. Government price support programs rewarded uniform commodity crops. The seed industry consolidated around varieties it could patent and sell. Within two generations, the accumulated genetic heritage of ten thousand years of farmer-led selection was replaced, across most of the world’s prime agricultural land, by a handful of commercially optimized varieties.

What Genetic Narrowing Actually Means

The numbers are stark enough to constitute an emergency, if we were paying attention. The Food and Agriculture Organization estimates that 75 percent of the world’s food plant varieties have been lost since 1900. Of the roughly 250,000 plant species that humans have eaten at some point in history, contemporary agriculture depends on fewer than 200, and just three — wheat, rice, and maize — provide more than half of all human caloric intake. The banana sold in every supermarket on earth is a single clone: the Cavendish, a variety that dominates global trade not because it is the most delicious or nutritious banana, but because it ships well and ripens predictably.

The Cavendish replaced the Gros Michel in the 1950s after Fusarium oxysporum — Panama disease — wiped out the previous monoculture. Right now, a new strain of Panama disease called Tropical Race 4 is spreading through Cavendish plantations across Asia, Africa, and Latin America. We are watching the same movie again. We know how it ends. The response of the global banana industry has largely been to hope TR4 stays contained long enough for someone to engineer a resistant Cavendish variant, rather than to diversify away from monoculture. This is not stupidity. It is the rational response to a market structure that rewards uniformity.

The problem is not just fungal pathogens. Genetic uniformity means uniform response to climate stress. The wheat varieties that dominate global production were bred for the temperature ranges and rainfall patterns of the twentieth century. As those parameters shift, the genetic toolkit available to breed adaptation into commercial crops is precisely the toolkit we spent the last century destroying. The wild relatives of wheat, the traditional landraces of rice, the heirloom tomato varieties that contain resistances to pathogens not yet evolved — these are the raw material for future plant breeding. We have been burning that library to heat our house.

The Political Economy of Seed Control

Crop diversity did not collapse by accident. It collapsed because a specific set of economic and political arrangements made it collapse. Plant variety protection laws, first enacted in the United States in 1970 and subsequently spread through international trade agreements, allowed seed companies to claim intellectual property rights over crop varieties. This created the incentive to develop proprietary varieties, sell them to farmers, and prevent farmers from saving seed. The farmer-breeder who maintained landrace diversity over centuries was replaced by the corporate plant scientist who develops a patentable F1 hybrid — a variety that, conveniently, does not breed true, requiring farmers to purchase new seed each season.

The consolidation of the seed industry has been one of the most dramatic in modern capitalism. In 1996, there were dozens of major seed companies globally. By the late 2020s, three companies — Bayer-Monsanto, Corteva, and ChemChina-Syngenta — controlled more than half of the world’s commercial seed supply. The bundling of seed with herbicide resistance (a plant is bred to tolerate a company’s herbicide, creating lock-in for both products) further concentrated power and further standardized the crop genetics in farmers’ fields.

This is not a story of villains. It is a story of incentives. Every step in this process was rational from the perspective of the actors involved. Farmers adopted higher-yield varieties because they needed to stay solvent in competitive commodity markets. Seed companies invested in proprietary varieties because intellectual property created return on R&D investment. Governments promoted uniform commodity crops because they stabilized food prices and simplified agricultural policy. The loss of diversity was a collective action problem of the largest possible scale — a cost distributed across the entire future of civilization, imposed by decisions that made economic sense to everyone who made them.

The Seed Vault Illusion

The standard technocratic response to this problem is the gene bank. The Svalbard Global Seed Vault, buried in permafrost on a Norwegian archipelago, contains nearly 1.4 million seed samples from around the world. Similar facilities exist in Fort Collins, Colorado; Wakayama, Japan; and dozens of other locations. The implicit argument is that we can outsource the maintenance of genetic diversity to cold storage — that the living, evolving, farmer-managed diversity of traditional agriculture can be replaced by frozen samples, available to be thawed and deployed when we need them.

This argument fails on several grounds. Seeds in cold storage do not evolve. A landrace maintained in a Mexican milpa for another generation accumulates new adaptations to shifting conditions; the same variety as a frozen sample from 1985 does not. The continuous co-evolution of crops with pathogens, insects, and climate — the process that makes traditional varieties resilient — stops at the moment of deposit. Gene banks preserve genetic sequences, not the adaptive intelligence embedded in living populations. They are like saving digital scans of books while burning the libraries. The information is not quite the same as the thing.

More practically, gene bank collections are incomplete, underfunded, and dependent on international political cooperation that has proven fragile. The Aleppo gene bank in Syria, one of the most important collections of dryland crop varieties in the world, was in a war zone for most of the 2010s. Its staff evacuated duplicate samples before the facility was damaged, which was genuinely heroic — but the episode illustrates how thin the margin is between the preserved diversity and its permanent loss.

What Comes After the Monoculture

The collapse of crop diversity is not a completed event. It is an ongoing process, and it is reversible at the margins — but only through deliberate policy choices that run against the current market logic. Those choices look like: seed sovereignty laws that protect farmers’ rights to save and exchange seed; public plant breeding programs that develop varieties for diverse smallholder conditions rather than commodity monoculture; agroecological subsidies that reward biodiversity rather than yield alone; and the patient, expensive work of supporting indigenous and traditional farming communities who are maintaining living diversity in the fields.

None of these are silver bullets. None of them can fully recover what has been lost. The landrace wheat varieties of pre-Green Revolution Anatolia, the traditional potato varieties that the Irish had abandoned before the Famine, the thousands of apple varieties that American orchards grew before the supermarket standardized on six — these are gone, or nearly gone, and frozen samples are a poor substitute for the living thing.

The argument I am making is simple: when you optimize a biological system entirely for short-term efficiency, you destroy its resilience. The Irish potato farmers of 1845 did not choose monoculture because they were foolish. They chose it because it was the rational response to their economic circumstances. We are making the same choice, at global scale, for the same reasons. The question is not whether the bill will arrive. It is whether we will have rebuilt enough diversity by the time it does to have anything left to grow.

Agricultural civilization has survived ten thousand years by maintaining optionality — dozens of crops, hundreds of varieties, farmers selecting for local conditions, preserving genetic hedges against futures they could not predict. We have spent the last century liquidating that optionality for yield. It is the most consequential trade we have ever made, and we made it without ever quite deciding to.