On Friday, the U.S. Department of Agriculture predicted that American corn farmers would have their worst season since 1995, thanks to the historic drought and sweltering heat that's seized much of the Midwest. Corn yields were expected to sink down to 123.4 bushels per acre. Global food prices are expected to jump.

To provide more context for the current drop, Stuart Staniford offers up this handy graph of U.S. corn yields going back to 1866:

The chart tells a nice, short story of U.S. agriculture — particularly corn, one of its most important crops. For most of the 19th century, American farmers were able to produce more and more food by planting on ever more acreage.* By the late 1800s, however, yields had stagnated. Policymakers got nervous, and the federal government undertook a series of initiatives to boost U.S. food production — irrigation and dam projects to bring farming to desert areas, railroads to transport food to cities.

The major gains, however, began in the 1920s and 1930s, when scientists began breeding hybrid strains of corn with bigger ears that could bunch more closely together in the field. New industrial fertilizers that could satisfy corn's voracious appetite for nitrogen were also developed. Tractors and other mechanized tools appeared. As the chart shows, yields began skyrocketing. Corn could now be grown in areas that were once unthinkable, like parts of the Great Plains. As Paul Roberts explains in The End of Food, "between 1930 and 1940, the number of bushels of corn per acre doubled, and then continued to rise each year."

As the chart above also shows, however, even after the explosion of corn production, the agricultural system is still sensitive to extreme weather events. The drought in 1988 caused corn yields to take a big hit. This year's drought has caused another steep drop.

So far, these lurches haven't been a fatal problem. For the past half century, droughts in the United States have actually been relatively short and infrequent, thanks to increased rainfall driven by natural ocean cycles. Globally, meanwhile, agricultural yields have been growing at a stable rate. Technology has helped farmers overcome nature. But recent modeling work by Aiguo Dai of the National Center for Atmospheric Research suggests that our luck might be starting to change. A combination of rising global temperatures and shifting natural ocean cycles will make drought in North America more common over the next 20 to 50 years. That could, potentially, cause even greater swings in U.S. food production.

As I wrote in a story for the paper on Wednesday, many scientists and seed companies like Monsanto are hoping to surmount this problem by engineering new drought-tolerant strains of corn that can thrive under hotter, drier conditions. The hope is that technology and better genetics can save the day, much as they did in the 1920s. But, as I explore in the piece, there are serious questions about whether drought-tolerant crops will be enough. Indeed, some experts, like William Moseley, have argued instead that the drought should cause the United States to rethink its dependence on corn, a lucrative and heavily-subsidized crop. Otherwise, greater volatility could be in our future.

Update: Just a quick correction, in the 19th century overall U.S. farm acreage grew, but individual farms were not growing in size, as originally noted. Thanks to bharshaw for pointing this out.