Phylogenetic studies that use many rapidly evolving genes provide great opportunity to address several questions about the domestication of plants and animals, including the timing, location and numbers of domestication events. An excellent example of such a study is Matsuoka et al.'s [1xA single domestication for maize shown by multilocus microsatellite genotyping. Matsuoka, Y.Y. et al. Proc. Natl. Acad. Sci. 2002; 99: 6080–6084Crossref | PubMed | Scopus (571)See all References][1] examination of the origin of maize (Zea mays) using microsatellite data. The land races of maize, domesticated from teosinte, exhibit striking morphological, karyotypic and genetic diversity. This diversity has led some researchers to the conclusion that maize, like many other domesticated plants and animals, had been cultivated several times independently. The results of Matsuoka et al., however, reject the hypothesis of multiple origins and instead show that maize arose via a single domestication event that most likely occurred ∼9000 years ago in Mexico.The authors sampled 193 maize accessions across its entire range prior to Columbus from northern Chile to southeastern Canada, 67 accessions from two subspecies of Mexican teosinte (Z. mays ssp. parviglumis and ssp. mexicana), as well as an outgroup. These were genotyped at 99 microsatellite loci scattered across the genome. All domesticated maize clustered into a monophyletic lineage within spp. parviglumis that had very strong bootstrap support. A contrasting pattern, where different cultivated ‘forms’ cluster with forms of wild relatives, has been found in several crops including cotton, rice and beans. Maize was probably first domesticated in Mexico because the most basal domesticated forms were those from the highlands of Mexico. The data also suggest that maize was first domesticated in the highlands and subsequently spread to the lowlands. This pattern is also consistent with the archaeological evidence.With a few exceptions, geographically similar forms of maize clustered with each other in the microsatellite tree. Population structure analyses identify three major clusters: US maize (with long and slender ears, particularly in the east), Andean maize (often with hand-grenade-shaped ears), and a group that included Mexican and South American forms. Despite the strong support for maize being derived solely from ssp. parviglumis, population structure analyses suggest that there has been modest yet measurable gene flow from ssp. mexicana into the maize of Mexican highlands.Provided the microsatellites have known mutation rates and evolve in accordance with a stepwise mutation model, they can be used to date the divergence of wild and cultivated forms. Based on data from the 33 (of 99 total) loci that fit the criteria, Mexican maize and teosinte diverged 9200 years before present (BP) with 95% confidence intervals spanning 5700 to 13 100 BP. For comparison, the oldest known fossil maize dates to 6250 BP but archaeological evidence suggests maize domestication could have begun as early as 10 000 BP.One major puzzle about maize domestication remains. The most primitive form of maize appears to be from the Mexican highlands, whereas the most likely ancestor of maize (ssp. parviglumis) is currently restricted to the lowlands. There are a number of possible explanations. Perhaps ssp. parviglumis had a larger range at the time when it was domesticated. Another possibility is that the inhabitants first transported ssp. parviglumis to the highlands to domesticate it. Solving this puzzle will most likely require a multidisciplinary effort involving archaeology, ecology, and molecular phylogenetic studies.