Cotton Story: The Crop That Fuelled Empires and Slavery Finally Reveals Where It Started

Farmers grow cotton across roughly 35 million hectares worldwide. Traders move it across commodity markets in New York and Zhengzhou. Textile mills in Bangladesh, India, and Vietnam spin it into yarn by the millions of tonnes. The plant behind all of it has become almost invisible—its biological origins, where it was first domesticated, by whom, and through what process, have remained, until now, genuinely unclear.

The species in question is Gossypium hirsutum, known as upland cotton, which accounts for around 90% of global cotton fibre production. It is not the only domesticated cotton. Gossypium barbadense, the long-staple Pima cotton prized by luxury textile makers, was independently domesticated in South America, near what is now Guayaquil in Ecuador, around 8,000 years ago. Two other species, Gossypium arboreum from the Indian subcontinent and Gossypium herbaceum from sub-Saharan Africa and the Arabian Peninsula, make up most of the rest. But G. hirsutum is overwhelmingly the dominant one. Every standard cotton T-shirt, every hospital bedsheet, every reel of cotton thread in a sewing box almost certainly derives from it. The question of where G. hirsutum was first brought under human cultivation is not an academic footnote. It is the origin story of one of the most economically significant plants in human history.

"Wild cotton plants are woody, multibranched shrubs or small trees, long-lived, with relatively sparse flowering and smaller flowers, fruits and seeds than under cultivation," Wendel told Reuters. "Members of some human groups must have taken an interest in the wild forms," he added, setting in motion the process of domestication from which the modern crop form arose over thousands of years of slow and gradual improvement.

Earlier studies, using older and less precise genetic tools, had pointed broadly toward the northern Yucatán Peninsula as the likely cradle of domestication. But the evidence was thin. Sampling of truly wild populations had been sparse, and the signal from the available data was too weak to pinpoint a specific region with confidence. The problem was compounded by a peculiarity of cotton's long history of cultivation: domesticated plants that escape into the wild can, over generations, shed enough of their cultivated characteristics to look phenotypically wild, resembling wild plants in their physical appearance, without actually being so. Telling a genuinely wild plant from a feral escapee in the field is fiendishly difficult. Only genomic analysis, reading the DNA directly, can do it reliably.

This is what made the new study's approach different. The research team collected 299 plants from across the full native range of wild G. hirsutum: 158 from the northern Yucatán Peninsula and 141 from coastal Florida, supplemented by previously sequenced samples from Caribbean islands including Puerto Rico and Guadeloupe. Each plant was genomically verified as genuinely wild before being included in the analysis. The result was the most complete and geographically expansive dataset of wild cotton genomes ever assembled, a resource that finally made it possible to ask, and answer, where this crop actually began.

What the data revealed was not simply a location. It was a portrait of a domestication process: gradual, cumulative, and far more complex than the sudden revolution that popular accounts of agricultural origins often imply. Early farmers recognised the plant's potential as a source of soft material; early weavers could spin its fibre by hand into cloth, fish nets, and rope. "Early farmers saw potential in this sprawling plant with hairy seeds as a source for soft materials," Grover notes. The transformation left clear genomic traces, all pointing unmistakably to one specific corner of the Yucatán coast.

Drive along the northern coast of Yucatán and you will not immediately recognise what you are looking at. The plants growing in the coastal scrubland are sprawling, multi-branched shrubs, chest-high in places, with small leaves and modest flowers that shade from pale yellow to pink as they age. The bolls they produce are small, and the fibre inside them is short and brownish, bearing no resemblance to the fat white tufts that spill from a commercial cotton field. These are not failed cotton plants. They are the original ones. And their DNA, it turns out, contains the most complete record of cotton's origins anywhere on earth.

The key concept here is genetic diversity, a measure of how much variation exists within a population's DNA. Think of it as biological memory. A population that has remained large, stable, and connected over thousands of years accumulates variation across generations; mutations arise, spread, and persist, building up a reservoir of genetic difference. A population that has been cut off from others, reduced in size, or pushed through a genetic bottleneck loses variation. A bottleneck is a severe reduction in population size that strips away much of the gene pool in a single generation, leaving survivors to carry only a fraction of what was there before. Diversity, once lost, does not come back easily.

By this measure, the wild cotton populations of northwestern Yucatán are extraordinary. Their nucleotide diversity, measured as the degree to which individual plants differ from one another at the level of their DNA sequence, is the highest recorded anywhere in the species' native range. Two plants drawn at random from this population will differ, on average, at nearly twice as many positions in their DNA as any two modern cultivars grown in a commercial cotton field today. Centuries of human selection, narrowing the gene pool in pursuit of whiter, longer, more abundant fibre, have left the crop genetically impoverished relative to its wild ancestors. "We know that domestication often leads to a loss of genetic diversity as early farmers were selecting for valuable traits, and then to further reductions as crop improvement intensified the selection pressure," Grover explains.

Every major form of analysis the researchers applied returned to the same address. Population structure analysis, a method for identifying genetically distinct groups within a species, placed the northwestern Yucatán plants closest to domesticated cotton. Phylogenetic analysis, which reconstructs evolutionary family trees from DNA, nested the domesticated gene pool within the northwestern Yucatán population. Plastome analysis told the same story. The plastome is the DNA of the chloroplast, the structure inside plant cells that converts sunlight into energy, inherited independently of the main genome and therefore a separate corroborating line of evidence. Even chemical profiling of wild plants along the Yucatán coast provided independent confirmation: a gradient in the plants' secondary chemistry runs from west to east, with western plants chemically closest to domesticated cotton. Every line of inquiry pointed to the same coast.

The picture that emerges from this convergence of evidence is precise enough to locate on a map. The domestication of upland cotton began in the northwestern corner of the Yucatán Peninsula, in the coastal scrublands that still exist there today. From that origin point, early domesticated forms spread gradually across Central America and the Caribbean, each dispersal accompanied by a loss of genetic diversity. Caribbean island populations, for instance, show only around one-quarter of the diversity found in northwestern Yucatán. The bottleneck of island colonisation is written permanently into their DNA.

Modern cultivars carry an additional complication. Roughly 13 to 14% of their genomes derive not from G. hirsutum at all, but from G. barbadense, the South American cotton species independently domesticated thousands of years earlier. As the two species spread across overlapping territories in the Caribbean, they interbred, intentionally and otherwise, and fragments of G. barbadense DNA were absorbed into the hirsutum gene pool. It is partly this genetic mixing that explains why modern cultivars do not sit neatly inside the wild Yucatán family tree. They are the product of a long, complicated history, one that runs well beyond genomics.

Cotton's spread across the world after the Spanish conquests of the 16th century was built on slavery, the exploitation of Indigenous peoples, and imperial expansion. The invention of the cotton gin at the end of the 18th century accelerated that history, making cotton farming enormously profitable and deepening the demand for enslaved labour across the American South. "Cotton has a complicated history, most notably its association with slavery, exploitation of Indigenous peoples and imperial expansion. But it is also an enduring crop, one that is woven into the lives of people worldwide," Grover observes.

What the study also found, and what carries perhaps the greatest consequence for the future, is that the domestication process itself left no single dramatic fingerprint. There are no major domestication genes, no single genetic switch that, when flipped, turned a wild plant into a crop. Instead, the researchers found thousands of small genetic changes distributed across the genome, each with a modest individual effect, accumulating slowly over millennia, consistent with how other major crops, including wheat and rice, were shaped by human selection. Cotton was coaxed into existence, incrementally, across thousands of years of patient human attention.

The fibres are single cells, each stretched to extraordinary length through millennia of selection, producing the fine, white staple the global textile industry depends on. "The fibers themselves are just single-celled seed hairs, but are among the most exaggerated and remarkable cells in plants," Wendel notes. "Research is showing that the process of domestication, of transforming these short, coarse and brownish fibers into the fine, white and superior textile we know today likely involves many genes operating in a complex symphony," Grover points out.

What the Coast Still Holds

Pull a cotton boll apart today and what you hold is the product of a 7,000-year conversation between farmers and a plant, carrying, on average, half the genetic diversity of the wild plants still growing on the Yucatán coast. That coastal scrubland is not a relic. It is a living seed bank, holding variation that thousands of years of human selection discarded in pursuit of whiter, longer, more abundant fibre, variation that plant breeders, facing a crop increasingly stressed by drought, heat, and pest resistance, may urgently need. Whether that scrubland remains intact long enough to be useful is a question no genome can answer.