"Ignoring climate projections at this stage will only result in the worst form of triage."
The headline is from the University of Washington news release on a study in Science, "Historical Warnings of Future Food Insecurity with Unprecedented Seasonal Heat" ($ub. req'd). The quote is the study's powerful final sentence. The release explains:
Rapidly warming climate is likely to seriously alter crop yields in the tropics and subtropics by the end of this century and, without adaptation, will leave half the world's population facing serious food shortages, new research shows ...
"The stresses on global food production from temperature alone are going to be huge, and that doesn't take into account water supplies stressed by the higher temperatures," said David Battisti, a University of Washington atmospheric sciences professor.
Worse, the study must also be considered a serious underestimate of likely impacts since, as is common in such analyses, they based their simulations on "the 'middle of the road' emission scenario, A1B." In 2100, A1B hits about 700 ppm with average global temperatures "only" about 3°C warmer than today. In fact, on our current emissions path, we are going to get much, much hotter.

Figure. "Histogram of summer (June, July, and August) averaged temperatures (blue) observed from 1900 to 2006 and (red) projected for 2090 for (A) France, (B) Ukraine, and (C) the Sahel. Temperature is plotted as the departure from the long-term (1900-2006) climatological mean (21). The data are normalized to represent 100 seasons in each histogram. In (A), for example, the hottest summer on record in France (2003) is 3.6°C above the long-term climatology. The average summer temperature in 2090 [assuming A1B] is projected to be 3.7°C greater than the long-term climatological average."
The results are still alarming: