The Evolution of Maize (Corn)
A couple of scientists working during the first part of the 20th century uncovered evidence that linked maize to what would seem to be a very unlikely parent: a Mexican grass called teosinte. Looking at the skinny ears of teosinte, with just a dozen kernels wrapped inside a stone-hard casing, it is hard to see how they could be the forerunners of corn cobs with their many rows of kernels. Teosinte is the common name of a wild grass that grows in several areas of Mexico and Central America. There are several closely related species of wild plants that have the common name of teosinte: Zea mays parviglumis, Zea mays huehuetenangensis, Zea mays mexicana, Zea luxurians, Zea diploperennis and Zea perennis.
But George Beadle (1903-1989) found in the early 1930s that maize and teosinte had very similar chromosomes. He made fertile hybrids between maize and teosinte that looked like intermediates between the two plants. Beadle concluded that the two plants were members of the same species, with maize being the domesticated form of teosinte.
However, his theory remained doubtful, because the differences between the two plants appeared to be too great to have evolved in just a few thousand years of domestication. So, Beadle sought ways to gather more evidence. He knew that one way to examine the parentage of two individuals was to cross them and then to cross their offspring and see how often the parental forms appeared. He crossed maize and teosinte, then crossed the hybrids and grew 50,000 plants. He obtained plants that resembled teosinte and maize at a frequency that indicated that just four or five genes controlled the major differences between the two plants.
Beadle’s results showed that maize and teosinte were closely related. But to pinpoint the geographic origins of maize, more definitive forensic techniques were needed. In order to trace maize’s paternity, botanists led by John Doebley selected more than 60 samples of teosinte from across its entire geographic range in and compared their DNA profile with all varieties of maize. They discovered that all maize was genetically most similar to a teosinte type from the tropical Central Balsas River Valley in southern Mexico, suggesting that this region was the 'cradle' of maize evolution. Furthermore, by calculating the genetic distance between modern maize and Balsas teosinte, they estimated that domestication occurred about 9,000 years ago.
In Xihuatoxtla, a natural rock shelter in the Rio Balsas valley (Guerrero, Mexico), they discovered an array of stone milling tools with maize residue on them. The oldest tools were found in a layer of deposits that were about 8,700 years old. This was the earliest physical evidence of maize use and it coincides very nicely with the time frame of maize domestication estimated from DNA analysis.
The domestication process must have occurred in many stages over a considerable length of time as many different, independent characteristics of the plant were modified.
 Matsuoka et al: A single domestication for maize shown by multilocus microsatellite genotyping in PNAS - 2002. See here.
 Ranere et al: The cultural and chronological context of early Holocene maize and squash domestication in the Central Balsas River Valley, Mexico in Proceedings of the National Academy of Sciences of the United States of America - 2009