Humanity has enjoyed an unusual streak of food surplus since the green revolution began in the mid-1960s. These trends sustained economic development and a significant reduction in global hunger and poverty. A sharp reversal is now possible, however, given strong economic growth in the world’s most populous countries and loss of suitable cropland.
People with rising incomes consume more meat and livestock products, which in turn requires more grain per unit of food produced. The rapid expansion of biofuel production only complicates the competition between food and fuel.
Moreover, yields of rice and wheat are running up against the genetic ceiling allowed by current varieties, and rates of yield increase are not sufficient to meet the demand for livestock feed, food and biofuels for the world’s 6.5 billion people. Without significant improvements, massive deforestation and environmental degradation will be inevitable in trying to feed the nine billion individuals who will be alive in 2050.
Debate is now raging over whether climate change will further reduce the world’s ability to feed itself. Estimating the long-term effects is critical to setting effective policies that ensure food security. Unfortunately, the answers differ. Much of this inconsistency arises because yield research conducted in greenhouses and on small plots, the current experimental methods, does not predict performance on commercial-scale fields; the conditions are just not comparable to production-scale farming. Without direct measurements under realistic growth conditions, we must resort to computer models or evaluations of historical data—and they show disparate results, too.
There is an urgent need to better quantify the impact of projected climate change on major crop yields. Funding for real-world experiments has been crashing, however. And linkages between models for climate change and crop production are relatively crude.
Policymakers depend on that work, but the models are only as good as the science behind them. The models’ predictions must be validated with real-world measures of how climate affects crops grown in actual agricultural ecosystems, over time and across regions. Without rigorous validation, models can mislead, as small errors expand into large ones.
Carbon sequestration in soil is a case in point. Models predict that soils will hold on to more carbon under so-called no-till farming practices, in which plant stalks and roots remaining after harvest are left to decompose. Yet recent studies based on direct measurement of soil have not confirmed any net improvement.
We cannot wait for perfect simulations; policy decisions must be made with imperfect knowledge. The danger, of course, is that poor policies built on erroneous models can waste billions of dollars. We must spend more on real-world research to improve the models so we can predict the impact of climate change. Only then can we decide whether the world can tolerate more crops for biofuels.
Note: This article was originally printed with the title, "Biofuels or Food?"