To some extent, the causes of agriculture are to be found in the thousands of years of hunting-gathering that immediately preceded the first agricultural societies. A critical interval may be the period between 22,000 and 8,000 years ago-a time of major climatic changes for much of the world. In Western Europe, population densities shifted as the herds of reindeer and horses that once supported many hunting bands moved northward as the glaciers became smaller. Some people moved with them, but others worked out subsistence strategies stressing plants, smaller game, and fish. Salmon became especially important in Europe as traps, drying racks, and other tools were developed to make salmon exploitation a reliable way to make a living. In Southwest Asia, parts of Africa, and parts of the Americas, some late Pleistocene and early post-Pleistocene (12,000-8,000 years ago) peoples began to eat more small game, fish, waterfowl, clams, wild cereals, and similar foods. Elsewhere, however, big-game hunting specializations (hunting large animals) persisted.

Where a shift to a smaller, more varied resource was made, technologies also changed. The bow and arrow and throwing stick were important innovations, and new tools were developed to did plants, trap wild fowl, and prepare and cook this broader diet. Small, simple, geometric stone tools predominated in many areas. The world about 12,000 years ago was relatively diverse culturally, as some groups remained big-game hunters while others took up fishing, intensive foraging, and other pursuits. Thus, a great diversity of plants and animals was being exploited with varying intensities and technologies in a wide range of climates. Out of this vast mixture of people, plants, animals, and places the first domesticates and farmer appeared.

But which of these groups became agriculturalists and why? Although all the people before about 10,000 years ago were hunter-gatherers, this term covers a wide range of economies. Much of the variability in hunter-gatherer adaptations seems to be linked to food storage. Anthropologist Alain Testart notes that some hunter-gatherers store large amounts of food, others do not, and that "storing hunter-gatherer societies exhibit three characteristics a sedentary (settled) way of life, a high population density, and the development of socioeconomic inequalities-which have been considered typical of agricultural societies". Thus, in looking for causes of agricultural economies, we must consider kinds of adaptations where storage is a potential factor. This relates directly to certain types of foods. Cereals store well, but many tubers do not. Sheep, goats, barnyard fowl, and other animals can also be considered a form of food storage, since one simply feeds them excess or unwanted foods until the need arises to eat the animals.

It is probably also significant, as researcher Kent Flannery notes, that the major cereal seed crops that supported the first farmers and remain the basis of modern economies, including wheat, barley, millet, and rice, appear to have derived from wild ancestors that were "third-choice" foods: plants that were usually more difficult to gather and process than other wild plants and thus were probably first eaten in quantity because people had to do so, not because they wanted to. On the other hand, most of these third-choice foods are easily storable, plentiful, easy to grow, and, as annuals, genetically adaptable.

Aspects of the shift to cereals and other resources may help explain how agriculture-once it was in its initial stages-changed human demographic patterns. In hunting-gathering societies, fertility rates are suppressed significantly simply by maternal mobility. Thus, those late Pleistocene groups that became less mobile, perhaps because they began to obtain food from salmon runs or wild cereal patches, might have experienced a rise in fertility rates. Also, a direct correlation exists between the amount of carbohydrates in the diet and fertility rates. Studies have shown that it is almost impossible for a woman to become pregnant until she has about 27.000 calories, or 20 to 25 percent of her body weight, stored as fat. Nursing a child requires about 1,000 calories a day, and in many hunter-gatherer societies, the rigors of mobility and their high-protein diet can mean that nursing itself prevents sufficient fat build-up for a successful subsequent pregnancy for about three years. But with the change to a high-carbohydrate, cereal-based diet and restricted mobility of sedentary life, fertility rates may well have risen rapidly.