For the foreseeable future, population will continue to grow, albeit it at a decreasing rate. Over the next half-century, the question seems not to be whether humans will produce enough food, but rather how that food will be produced. The human population is passing through what E.O. Wilson has called the bottleneck, and by the middle of the twenty-first century, it is likely to level off at somewhere between 8 and 9 billion people.

Overall food production can be achieved in just a few ways:

• Distribute food more equitably by curtailing over-consumption and reducing the production of meat
• Increase crop yields
• Increase the land area under cultivation

Each of these broad strategies involves a lot of possible specific cases and a number of complicated trade-offs. The story of soybeans in Brazil — particularly in the huge, interior state of Mato Grosso — illustrates several of the complications associated with the second and third options.

I took these photographs of Cargill’s riverfront terminal just downstream from Porto Velho, Rondônia in 2003. Much of the soy grown in the center-west portion of the country is brought by road to this break-in-bulk point, where it is transfered to barges that can take it all the way to Belem, at the mouth of the Amazon River, for export.

For many years, agricultural production in Brazil increased slowly, if at all, and sometimes not fast enough to keep up with domestic population growth. What little increase did occur was strictly the result of increases in the amount of land being cultivated. Even though considerable efforts were made to increase yields, improvements in technology did little more than offset the poor quality of the new lands being cultivated.

The twentieth-century experience of Brazil is hardly surprising; humans farm about 1/8 of the earth’s land surface and almost by definition this is the most productive 1/8. Any new areas brought into production are likely to be marginal lands in both senses of the word: in peripheral locations relative to existing human settlement and of lower quality relative to already-settled lands.

By the close of the twentieth century, however, something clearly had changed, as Brazil’s agricultural output — particularly of soybeans — began to challenge the role of the United States as the dominant producer in the Western hemisphere. Reporting for the radio program Living on Earth, Bruce Gellerman has described this transition beautifully in Magic Seeds and the Miracle Crop. (His report is available as an mp3 and as text with some excellent photographs.)

The report describes how Mato Grosso has become such a large and still-growing producer of soybeans, despite the unsuitable soil conditions. It then goes on to describe the consequences associated with such success: increased reliance on pesticides, the tendency of crop pests to develop pesticide resistance, and the great loss of habitat in the savanna, known in Brazil as cerrado, or “closed” for its traditional inaccessibility.

Brazil is advancing on its frontier just as the United States did more than a century ago. A decade ago, I wrote  comparison of the two frontier experiences: that of the United States in the 19th-century West and that of Brazil in the 20th-century Amazon. According to Gellerman’s report, the process continues in the 21st-century cerrado, but with more than one biome at stake and with the potential for much more substantial clearing. With the techniques currently being employed, the area remaining to be cultivated in Brazil might be greater than the area currently cultivated in the entire United States.