Researchers at the University of Illinois at Urbana-Champaign have published the results of a study aimed at identifying the amount of marginal land available worldwide for biofuel crop production. Partially funded by BP’s Energy Biosciences Institute and co-sponsored by the University of California, Berkeley and the Lawrence Berkeley National Laboratory, the study focused on determining what types of land are suitable for biofuel crops, where that land is located and what the current land cover is, according to U of I civil and environmental engineering professor Ximing Cai.

“Many, many studies focus on the productivity of biofuel crops,” Cai said. “But if we don’t have enough land anyway, then we could not answer the question about the real potential of bioenergy in the future.”

Because the criteria used to define marginal land vary vastly from study to study, researchers found it extremely difficult to precisely estimate the amount of marginal land suitable for biofuel crop production in various countries. Because of these issues, the researchers settled on presenting a global estimate based on physical characteristics of land using existing global databases.

Cai and his team used Fuzzy Logic Modeling to treat the uncertainties in land productivity assessments in the databases. Then, the team estimated the marginal land available for biofuel production in Africa, China, Europe, India, South America and the U.S. using several scenarios. The first considers only mixed crop and natural vegetation land. The second scenario adds marginal cropland. The third also considers using low-input high-diversity grassland for biofuel production, although the research team said this scenario is likely not realistic considering those lands are already used for pasture and cropping. The final scenario discounts the pasture land that was included in the third scenario and adds regular marginal land as well as regular and marginal grassland, savanna and shrubland.

In each scenario considered by the research team, Africa and South America had the most available land for biofuel production. According to the first scenario, which uses only marginal mixed crop and natural vegetation land, approximately 320 million hectares of land are available for biofuel production globally. South America had approximately one-third of the total amount of available land in that scenario while the U.S. had 43 million hectares of available land. The third scenario, which was deemed unrealistic by the researchers, resulted in the most available land for biofuel production worldwide—1,411 million hectares. In that scenario, Africa had 481 million hectares of available land while the U.S. had 127 million hectares of available land.

The results of this study indicate that biofuels produced on abandoned or degraded cropland could provide up to 52 percent of the world's liquid fuel consumption. Additionally, according to the researchers, planting second-generation feedstocks on marginal croplands and low-input, high-diversity prairie on marginal grasslands could be utilized to ultimately fulfill up to 58 percent of the current global liquid fuel consumption, without affecting pasture land and without compromising land currently being used for conventional crops.

Cai cautioned against taking the study’s results too literally, however, and said they are meant only to provide a basis for future research. “The purpose of our study was, we just wanted to get started on the land availability assessment,” he said. “We don’t want to say we provide some accurate numbers, although we do provide some numbers. We just hope other studies can [find] some way to verify our data or use our data for their planning purposes. For example, agricultural economists could use the dataset to do some research with the impact of institutions, community acceptance and so on, or some impact on the market.”

Cai said he has received a large number of requests for data from the study already. The study, “Land Availability for Biofuel Production,” is available at http://pubs.acs.org/doi/abs/10.1021/es103338e.