American Antiquity, 19(4): 409-410 (Apr 1954)
New Evidence on the Origin and Ancestry of Maize
Paul C. Mangelsdorf

New evidence concerning the origin of maize—a question which has puzzled botanists for more than a century—has made it possible to reach several rather definite conclusions:

  1. Maize is undoubtedly an American plant.
  2. Maize undoubtedly had at least one center of origin in Middle America.
  3. The ancestor of maize is maize.
  4. The ancestor of maize is a form of pod corn, but perhaps not the extreme type of pod corn known today. The ancestor was certainly a popcorn.
  5. Sometime in its history maize hybridized with Tripsacum or teosinte or both to produce radically new types which comprise the majority of modern maize varieties of North America.

The evidence on which these rather sweeping conclusions rest comes from three fields—botany, archaeology, and genetics.

Substantially conclusive evidence of the American origin of maize was obtained from fossil pollen grains discovered at a depth of more than 200 feet below Mexico City. These fossil grains were found in drill cores studied by Paul Sears of Yale University and Kathryn Clisby of Oberlin College. First thought to be pollen grains of teosinte they have now been identified by Elso Barghoorn of Harvard University and Margaret Wolfe of Radcliffe College as fossil pollen grains of maize. In making this identification Barghoorn and Wolfe compared the fossil grains with modern pollen grains of maize and its two relatives, teosinte and Tripsacum, not only in size but also in the ratio of the diameter of the pore to the diameter of the entire pollen grain. In this characteristic the fossil pollen is quite similar to modern maize and differs from both teosinte and Tripsacum. On the basis of currently accepted glacial chronology the fossil maize pollen is at least 60,000 years old. It thus antedates by many thousands of years the earliest evidence of agriculture or of man himself in this hemisphere. It seems probable that the fossil pollen is that of wild maize which once grew in the Valley of Mexico.

New archaeological evidence bearing upon the problem comes from caves in Mexico, New Mexico and Arizona.

In La Perra cave in the state of Tamaulipas, Mexico, Richard MacNeish of the National Museum of Canada turned up small cobs of a primitive maize dated by Libby's radiocarbon determinations of associated vegetal remains at 4445±180 years. My colleague, Galinat, and I have studied the botanical characteristics of these early La Perra cobs and have concluded that they represent the prototype of a small-eared race of popcorn, Nal-tel, which is well illustrated on Mexican funerary urns and which is still grown by the Indians in Yucatan and Campeche. The La Perra maize appears to be an agricultural type in the early stages of domestication.

Even more primitive cobs have been identified in archaeological material dug up in 1950 by Herbert Dick, then of the University of Colorado Museum. This maize comes from Bat Cave in New Mexico and is dated at 5605±290 years by Libby's radiocarbon estimates of associated charcoal. Dick, in an earlier expedition in 1948, had found in Bat Cave the oldest and most primitive corn then known. In the more recent expedition he turned up still smaller and more ancient specimens.

The small cobs of the Bat Cave maize are scarcely larger than a one-cent piece. They are extraordinarily well preserved and it is possible to study their botanical characteristics in detail. The most primitive cobs are characterized by a very slender rachis, glabrous cupules, long rachillae and tiny kernels, approximately fifty in number, which are almost completely enclosed in long soft glumes. The long rachillae and long, soft glumes are characteristics of pod corn. Remains of several kernels were found along with pieces of tassel, husks and pollen grains. From the combined vegetal remains it has been possible to make a tentative reconstruction of the ancient maize plant. Apparently this primitive maize had a short slender stalk not more than a few feet in height bearing a few kernels at the base of the tassel and a miniature ear immediately below the tassel. The ears were enclosed in relatively long husks which spread open at maturity. A plant of this kind probably could have existed in nature, and, although this earliest Bat Cave maize may have been cultivated, it was not far removed in its botanical characteristics from wild maize.

Although both the Bat Cave maize and the La Perra maize in Mexico are primitive, they differ in a number of botanical characteristics to the extent that we regard them as two types which were domesticated separately from geographically distinct races: one adapted to the lowlands and the other to the highlands. We suspect that there may have been additional races of wild maize in Central America and perhaps in South America as well. The recent discovery of wild beans in both Middle and South America lends some support to this suggestion.

The primitive Bat Cave maize proves beyond a reasonable doubt that the ancestor of maize was maize and not, as some nineteenth century botanists have supposed, the related grass, teosinte.

Yet teosinte (or another related grass, Tripsacum) has clearly played an important part in the evolution of maize. Prehistoric cobs from other caves show evidence of the contamination of domesticated maize by teosinte. This evidence comes from Cebollita Cave in New Mexico, from two caves in Arizona, and a cave in Chihuahua, Mexico. These were explored respectively by Reynold Ruppe of the Peabody Museum of Harvard University, Lloyd M. Pierson of the National Park Service, and Robert H. Lister of the University of Colorado Museum. In the material from each of these caves there are specimens of cobs which show strong evidence of contamination by teosinte. Furthermore, it is possible to match some of the archaeological specimens almost exactly with synthetic types produced by actually hybridizing maize and teosinte. The hybridization of maize and teosinte may have been a comparatively recent event in the history of maize under domestication. In the material from Cebollita Cave the early maize is all "pure" maize while the more recent material shows strong evidence of contamination.

Although there are many questions yet to be answered, the main features of corn's origin and evolution are now reasonably clear. The American Indians apparently domesticated maize wherever they found it. Domesticated varieties originating from distinct geographical races soon hybridized to create new and more productive types. Still later corn crossed with teosinte. To new races resulting from this cross teosinte contributed not only resistance to drought and diseases and the structural strength needed for the development of large plants and ears, but also mutability.

All of these factors combined to speed the evolution of maize to a point beyond that of perhaps any other cultivated plant. In 6000 years, or less, a small wild grass bearing tiny ears no larger than a modern strawberry has evolved into one of the world's most productive cereals. The evolution of maize has been truly explosive. Most of the factors involved in this extremely rapid evolution have been largely accidental. There is little evidence that man practiced artificial selection in maize in the early stages of domestication, or if he did, no evidence that he succeeded. His role was primarily one of creating opportunities for hybridization between races of maize and between maize and its wild relatives. But when man finally did begin to practice selection, he had a rich diversity at his disposal from which it was possible to choose a combination of characteristics that have made maize one of the most efficient of all cereals as a producer of foodstuffs.

PAUL C. MANGELSDORF
Botanical Museum
Harvard University
Cambridge, Massachusetts