Unravelling cotton’s domestication

A guest post from Ola T. Westengen, Coordinator of the Svalbard Global Seed Vault in Norway.

I am one of these nutcases that continues to be intrigued by the kind of research questions pioneered by Vavilov and I was flattered when Jeremy asked me to write a short guest post on the domestication and spread of cotton. He asked because Vavilov wrote about a “pitiful” cotton species that he encountered in the lowest agroecological zone of the Pamirs; ^[1] Jeremy then found out that I had spent time studying the diversity and origin of cotton. ^[2] This, however, was on another continent — South America — but I’ll have a go anyway. So what does New World cotton have to do with the miserable Pamir cotton described by Vavilov?

Here is a blitz tour of cotton domestication: In the annals of crop domestication cotton has a unique position because four species of the genus Gossypium were domesticated in four independent events that focused on the same trait in each case. Four different groups of pre-historic people recognized that the elongated epidermal seed trichomes of two Old World cottons (Gossypium herbaceum and G. arboreum) and two New World cottons (G. hirsutum and G. barbadense) were useful for spinning and weaving.

Archaeological evidence of the remains of cotton in association with human settlements goes back to the 8th millennium BP for the Greater Indus area in the Old World, and to layers dated to 6400–5000 years BP for the Zaña river valley in present day northwest Peru. For the Old World species, domestication studies propose that G. herbaceum has its domestication centre in Arabia, and G. arboreum in the Indus Valley of Pakistan. The New World species have received more scientific attention due to their present day economic importance and for G. hirsutum the Yucatan Peninsula in Mexico is the probable domestication center, whereas for G. barbadense the areas flanking the Guayaquil gulf of southern Ecuador and northern Peru is the most probable origin.

Thus, the origin of the cotton found by Vavilov in Pamir was apparently independent from that of the cotton in the Americas. Or was it? At a deeper level it was not, and at the risk of losing a few readers I’ll also give a short brief on the cytogenetics of the four cottons.

There are eight diploid genome types ^[3] within the Gossypium genus. The genome types are named with the letters A-G and K for convenience. Then there is one tetraploid genome type ^[4] that is a hybrid of the A and the D genomes. The tetraploid species (there are five of them) all comes from the New World and the two New World domesticates described above both are both tetraploids. Now, here is the intriguing part: the A genome originated in the Old World whereas the D genome originated in the New World. Thus, the hybridization leading to the tetraploids of the Americas seems to have been transoceanic. Even more intriguing, the most probable donor of the A genome is the Old World domesticate G. herbaceum, the very same “pitiful” species Vavilov observed in the Pamirs! This naturally resulted in speculations about transoceanic voyages by people carrying G. herbaceum from the Old World followed by a hybridization event with a local D genome donor that then resulted in the high yielding New World tetraploids. Alas, recent molecular work has shown that the hybridization took place some 1.5 million years ago. Nature did all the prebreeding in this polyploidy-creating event.

Perhaps this diagram ^[5] will help?

Nevertheless, the cottons brought back to the Old World as part of the Columbian exchange were superior to the pitiful diploids people over there had been wrapping themselves in. Today almost all the cotton cultivated around the world is of the species G. hirsutum with humble origins in Yucatan. I see from the production statistics that Uzbekistan is the second biggest cotton producer after the US in the world. I wonder how much G. herbaceum someone going in Vavilov’s footsteps would find today?

Notes:

1. Photo of herbarium specimen from the Natural History Museum, London. [↩]
2. I had my cotton adventure collecting landraces along the cost of Peru, crossing over the Andes and down in the Amazon. It was a true adventure, involving trips to remote villages on muleback and visits to dusty archeological excavation sites. The joys of Vavilov-style traveling! And if you want to see the references for the claims in this late night post try the paper we got out of the trip: Ola T. Westengen, Zósimo Huamán, Manfred Heun (2004). Genetic diversity and geographic pattern in early South American cotton domestication Theoretical and Applied Genetics, 110 (2), 392-402 DOI: 10.1007/s00122-004-1850-2. [↩]
3. Diploids contain two copies of each chromosome; this is the normal state of affairs for most flowering plants. [↩]
4. Tetraploids contain four copies of each chromosome. Many domesticated species are tetraploid, or have even higher ploidy numbers. [↩]
5. From Wendel & Cronin 2003. [↩]

Tagged as: Cotton Domestication