Study: Global feedstock yields are overestimated

By Erin Voegele | January 12, 2009
Web exclusive posted Jan. 16, 2009, at 4:22 p.m. CST

A recent study has found that global yields of most biofuels crops have been overestimated by 100 percent to 150 percent or more, suggesting that many countries need to reset their expectations of agricultural biofuels to a more realistic level. The study was led by Matt Johnston and Tracey Holloway with the University of Wisconsin-Madison Nelson Institute for Environmental Studies and Jon Foley with the University of Minnesota. Agricultural data from nearly 240 countries was used to calculate the potential yields of 20 different biofuels feedstocks, including barley, cassava, maize, potato, price, sorghum, sugarbeet, sugarcane, sweet potato, wheat, castor, coconut, cotton, mustard, oil palm, peanut, rapeseed, sesame, soybean and sunflower.

The study indicated that the potential for biofuels production in both developing and developed countries has often been exaggerated because current yield estimates are generally based on data sourced from the United States and Europe, which doesn't account for local differences in climate, soils, technology and other factors that can influence agricultural yields.

By offering an analysis of detailed, regional yield data that encompasses this variability, the researchers hope to empower countries to make wiser choices on whether to invest in ethanol or biodiesel, which crops to plant, and how to make the best use of exiting farmland. In the past, Johnston said, policymakers, companies and farmers have partially based decisions about biofuels on yield tables, which make simple side-by-side comparisons of the fuel yield per unit of land for various crops. For example, a yield table might compare the amount of ethanol one hectare of sugarcane can generate, compared to one hectare of corn. He said the problem with yield tables is they often reflect the best case scenario. "They might pick Brazilian sugarcane, versus rapeseed from France, versus maize from the United States," Johnston said. "Those aren't really fair comparisons because you are really cherry picking the best information from the best results from everywhere in the world, and putting them side by side. That's not very realistic for most people or places in the world."

To get a more realistic picture of the potential yields of these crops, the researchers used a global agricultural database, which was developed at the UW-Madison's Nelson Institute Center for Sustainability and the Global Environment (SAGE). The database, referred to as the M3 database, provides actual yields for 175 crops, circa the year 2000, at a resolution of roughly five miles by five miles across the entire globe. According to Johnston, the M3 database contains more than 22,000 unique census information datapoints from around the world that was sourced from the USDA and similar organization from around the world. "Basically, it's a compilation of agricultural census information from around the world," he said.

Johnston sourced data from the M3 database on the 20 biofuels crops he's studying to calculate and map the amount of biofuel that could be produced per hectare in every possible country by crop combination. He then computed a global average yield for each of the 20 feedstocks, as well as the average yield of each in both developed and developing nations as a whole.

"What we wanted to do was calculate yield potential," Johnston said. You can't say that just because you get a certain corn yield in Iowa that you can realize that yield everywhere else in the world, added. "That's just not true. Crops grow better in certain regions because of climate or water or soil conditions," Johnston said.

The next step is to compare the yields of biofuel feedstocks in areas with similar climates, and then study how differences in management practices may be contributing to production gaps. The goal is to help supply countries and farmers with the best data possible so they can make the most of existing farmland and better balance the investment in biofuels against other needs.

The basic idea will be to compare apples to apples. "We group all the points in the world that have similar water, soil, and climate suitability," Johnston said. These points are divided into 100 unique bins, or unique groups, with each one being considered a different agricultural suitability. "The idea is you can compare different places around the world that share similar climate conditions, and then compare the yields," he said. "Then you can say within this bin, we should be able to achieve what some of the top countries are - what some of the top plots are." The research should give farmers around the world a tool to help compare their plots of land with other plots of land around the globe that share similar climate, soil and water conditions. "Yield potential is what we are trying to calculate," Johnston said. "We can use this tool and the result of this study to optimize efforts [to improve yields]."

The results of the research have been published in an open access journal, called Environmental Research Letters. The full research article, titled "Resetting global expectations from agricultural biofuels," can be accessed on the Environmental Research Letters Web site. In addition, supplemental materials and access to data used to complete the research is available on the SAGE Web site.