interhalder says he and his colleagues are just beginning to make some headway on the answers. Dolores Piperno and Deborah Pearsall’s The Origins of Agriculture in the Lowland Neotropics, published in 1998 by Academic Press, adopted a behavioral ecology approach pioneered by Winterhalder and his colleagues, with implications for contemporary issues such as the preservation of seed diversity and natural habitats. The book inspired Winterhalder and Doug Kennett, an archeologist at the University of Oregon, to sign an agreement with Smithsonian Institution Press to produce a volume titled Foraging Theory and the Origins of Agriculture, which will include contributions from archaeologists around the world who are working with similar models.

The models, which are analogous to those economists use to understand the behavior of consumers in a market economy, attempt to account for part of the bewildering diversity of the choices made by individuals who live instead within ecosystems. Within that diversity, Winterhalder and others are finding common motives, common themes. One of them is avoiding risk, which he says is not just a motive of hunter-gatherers sharing food. Even when a population settles into agriculture, risk goes with the territory.

Among farmers in the Andes of Peru, Winterhalder found that a farm family might scatter a dozen or more fields hither and yon over high altitude ridges and valleys. This is not because Andean farmers are inefficient, or because they especially enjoy tramping up and down the mountain slopes. Winterhalder’s team gathered data on over 700 agricultural plots managed over a period of two years by 20 families in the Andes.

Using a computer-based geographic information system, which correlates various layers of information to positions on the landscape, the team plotted the paths between these fields, elevations and distances, and then calculated such things as the time it took to walk, the number of trips to the fields, and the amount of seed and fertilizer the farmers carried.

"When we calculate all of this," Winterhalder says, "it turns out that the cost to net production of this scattering is about seven percent, which is quite a lot lower than economists had actually estimated."

interhalder then posed the question: "What if this family had grown everything in just one of the spots where one of its fields was located. What would have been their production that year? And what would have been the variability, especially their chance or risk of failing to produce enough?" The variability, he found, was enormous. For each position in the landscape, weather, sunlight, soil conditions, pests, crop diseases, and other factors varied significantly from one year to the next, and the effects could be drastic.

"On any given year, there was a high probability that a family farming at only one or two locations would not produce enough food to meet their needs," Winterhalder says. "But if you scatter that production over twelve or thirteen fields, the reliability of producing enough to feed your family approaches one hundred percent."

American farmers typically don’t scatter their fields because the federal government offers crop insurance and other ways to avoid the risk of crop failure. The Incas, with their sophisticated system of central grain storage and distribution, also probably had no need for field scattering, Winterhalder says. But the common-sense idea behind field scattering remains ingrained in our behavior, even today. From childhood, we’re warned not to put all of our eggs in one basket. Gamblers hedge their bets. Investors balance their portfolios.

This kind of decision-making, Winterhalder says, may have originated with bands of hunter-gatherers sharing food whose sources were asynchronous — that is, the sources and the success of individual hunters and gatherers varied from day to day. In fact, the typical hunter-gatherer band of 25 to 45 people, of whom some 6 or 8 would have been productive hunters, was, according to Winterhalder’s mathematical analysis, just large enough to ensure that everyone got fed, even if most of the hunters came up empty some days.

"Even if there’s only a small amount of asynchrony in the sources you’re pooling, you don’t need very many folks to get a pretty large advantage," Winterhalder says. "Six, seven, or eight hunter-gatherers give you that advantage."