Sam Cox       December 2000 

Domestication origins are often difficult to locate, since most domestications of modern crops took place several thousand years ago, and archaeological evidence is difficult to come by, or difficult to interpret when found. In addition, crops have undergone constant change since domestication, thus, an ancient domesticated crop is often difficult to distinguish from its wild progenitor in fossil records until long after domestication, when the crop has attained significant morphological uniqueness. Archaeological remains of crops are infrequent, and strongly associated with certain climates which preserve plant material better. For these reasons, it is difficult to conclude definitively that any crop had more than one domestication center since the alternative, one domestication with subsequent diffusion, is difficult to disprove. Nevertheless, strong evidence for multiple domestication of some crops exists, and one can conclude with only minor doubt that some crops did indeed undergo domestication in different areas at different times. 

BEAN

Strong evidence supports the theory that beans have been independently domesticated one or more times. Four species of beans have been domesticated from the thirty or so wild species of South and Central America. These are Phaseolus vulgaris, P. lunatis, P. acutifolius and P. coccineus, the first two being the prime suspects of multiple domestications (Kaplan, 1965; Harlan, 1971; Heiser, 1965; Evans, 1976)). 

Examinations of the wild Phaseolus species form which domestications likely occurred revealed that these wild populations were often morphologically distinct, geographically isolated, and in some cases non-interfertile, as in beans from Mesoamerica and South America (Gepts, 1993). Wild beans are members of the species P. aborigineus. Grouping all these populations under one species is technically questionable, but has been long the standard, and is unlikely to change under new evidence (Sauer, 1993). 

Strong evidence for multiple domestication of wild progenitors involves a seed storage protein called phaseolin, which has eight different genetically controlled types among wild populations (Gepts, 1993). It is significant that among domesticates, six of these protein types are present. Some take this as a sign of multiple domestication events (Harlan, 1992). It is unlikely, given the geographical isolation of the wild populations, that one domestication incorporated that many protein types, and is only slightly less unlikely that one people domesticated multiple types. The dominant presence of some of these protein types largely rules out introgression from wild species since introgressed genes would be present only in isolated instances instead of widely in the general population, as in the rare Columbian B-type protein (Blumler, 1992). Additionally, beans only rarely cross pollinate, so introgression from wild species should not be a significant source of gene flow. The majority of bean types contain either the S-type, found in beans from Mesoamerica, or the T-type, found in beans from South America (Sauer, 1993). 

Mulitple domestication theory of beans received a boost in 1996 with the discovery of differential presence of the fin gene between accessions from South America and Mesoamerica. The fin gene causes a dwarf growth habit, as opposed to typical climbing habit. It reduces the number of nodes, time to maturity, and height of the plant by transforming apical vegetative meristems into terminal reproductive meristems (Koinange, et al. 1996). Traditional bean cultivation in Mexico involved beans climbing up corn stalks. The absence of corn, at least early on, in Peruvian culture would make a bush bean preferable for cultivation. 

Arhcaeological evidence also supports at least two independent domestications of beans. P. vulgaris was present in both Peru and Mexico by 5000 BC. P. lunatis was present in Peru by 6500 BC and in Mexico by 800 AD. The presence of  P. vulgaris in both locations at the same time strongly supports independent domestication. The later appearance of P. lunatis in Mexico is not as convincing for independant domestication, since certainly 6000 years is long enough for diffusion to take place. However, beans found in archaeological sites from South America have consistently larger seeds than those found in Mexico. This difference in seed size suggests different domesticates. 

Isozyme and morphological traits also support independent gene pools stemming from independent domestications (Gepts, 1993). 

In summary, at least two independent domestications of beans have seem to have taken place; one in South America and one in Central America. Possibly more than one domestication took place in South America. The evidence for this includes differences in isozymes, phaseolin protein types, morphology, geography of wild ancestors and the presence of the fin gene. Archaeological remains also support independent domestication. 

COTTON

An important fiber crop for several civilizations, cotton was undoubtedly domesticated multiple times. Very little argument on this point can be made since domesticated forms were found on both sides of the Atlantic much earlier than any known trans-Atlantic travel. However, independent domestication has even been suggested in the New World, based on several lines of evidence (Pickersgill, 1977). Altogether, cotton has likely been domesticated four times around the world (Sauer, 1993). 

There are four domesticated species of cotton. Prior to the relatively recent expansion of cotton production into temperate zones, all cultivars were perrenial shrubs or trees (Harlan, 1992). Only species possessing the A genome produce filamentous fibers used for textiles. 

East of the Atlantic, the diploid (AA) Gossypium herbaceum was most likely domesticated in Ethiopia. The exact date is not known, but Theophrastus recorded seeing a "wool bearing tree", which was probably cotton, in 350 B.C. G. arboreum, another AA diploid, seems to have been domesticated in India, where the current center of diversity is. This cotton was more widely used than G. herbaceum in the Old World prior to trans-Atlantic travel. 

The New World has four wild tetraploid species (AADD), two of which have been domesticated. Molecular evidence shows that all of the tetraploid cottons contain the same A genome, suggesting that the A genome was incorporated into cotton a very long time ago, perhaps 1-2 million years ago, with subsequent speciation into four distinct groups (Gepts, 1993). Diploid New World cottons do not contain the A genome (Harlan, 1992). The A genome is present, in more or less the same form, on both sides of the Atlantic (Sauer, 1993). This has caused some consternation about the source of the A genome, and possible transport across the Atlantic. The most plausible explanation is that the A genome is very ancient, and probably occupied a range on the ancient continent of Pangea over much of current South America, Africa and India (Harlan, 1992). The immediate ancestor of either cotton species is not definitively known, and may be extinct (Gepts, 1993) 

Evidence of multiple New World domestication is primarily archeaological. G. barbadense was domesticated in South America. Remains of this cotton were found in a site dated to 2500 B.C. G. hirsutem was probably domesticated in Mexico, since definitive remains of domesticated cotton have been found at a site dated to 3500 B.C. Both domestication centers are centers of diversity for their respective species (Gepts, 1993). The early presence of both domesticated species in geographically isolated areas strongly suggests independent domestication. Although the two species are morphologically similar, they are easily distinguished by isozyme alleles (Gepts, 1993). 

Given the evidence, it can safely be concluded that cotton was domesticated four times: in India, Africa, Mexico and South America. 

References

Blumler MA. 1992. Independent inventionism and recent genetic evidence on plant domestication. Economic Botany. 46(1):98-111. 

Evans AM. 1976. Beans. In: Evolution of Crop Plants. Edited by: NW Simmonds, et al. Longman; London. pp 168-172 

Gepts P. The use of molecular and biochemical markers in crop evolution studies. In: Evolutionary Biology, V27.1993 Edited by: MK Hecht, et al. Plenum Press, NY, NY. 

Harlan JR. 1992. Crops and Man, 2ed. Ameican Society of Agronomy, Inc.; Madison, WI. 

Harlan JR. 1971. Agricultural origins: Centers and Noncenters. Science 174:468-474. 

Heiser CB. 1965. Cultivated plants and cultural diffusion in nuclear America. Am. Anthropl. 67:930-949. 

Kaplan L. 1965. Archaeology and domestication in American Phaseolus. Econ. Bot. 19:358-368. 

Koinange EMK, Singh SP, Gepts P. 1996. Genetic control of the domestication syndrome in common bean. Crop Sci.36:1037-1045. 

Sauer JD. 1993. Historical Geography of Crop Plants. CRC Press; Boca Raton, FL. pp73-80.