Yielding Questions

What role does yield play in the carbon debt from land use change?


Gretchen Daily

Link to a recap of the workshop on life-cycle greenhouse gas analysis for the proposed revisions to the National Renewable Fuels Standard program, including audio and slide presentations from Sheehan, Fargione and some 20 other participants. More >>

This past summer, noted biofuel experts John Sheehan and Joe Fargione took part in a spirited debate over greenhouse gas emissions from land use change, with a focus on agricultural yield improvements. Hosted by the Environmental Protection Agency in Washington, D.C., the workshop helped inform the proposed revisions to the Renewable Fuel Standards (RFS) program.

“I wanted the EPA to acknowledge that if background yields improve faster than the demand for land and food, those yields can dramatically lower land use impacts,” Sheehan says, citing biotechnology as a potential player.

But Fargione is concerned that agricultural demand will outstrip yield increases, causing continued expansion of agriculture into remaining natural areas. What follows are excerpts from the in-depth exchange between the two scientists.


Scientific Program Coordinator, Institute on the Environment

“Let’s look at the 16 billion gallons of ethanol that RFS would like to have occur by 2022. An assumption is that all this land is not marginal land but prime agricultural land. If no yield improvement occurs after 2007, the model suggests you are going to see a huge demand for land for agriculture.

It probably doesn’t make sense to assume no yield improvement in the future. It will occur. And if you listen to Monsanto … it will be more dramatic than the historical trends. Even given growth in population and food demand, our model suggests there is a period out there around 2035 when the amount of land we need in agriculture could actually begin to decline. If you look at the annual emissions from carbon that are occurring from year to year, given historical yield improvements, the carbon debt from land clearing is much, much smaller and you actually pay it back sooner.

In a constant yield future, there’s a huge carbon debt. In a historical yield case, you could see a payoff in 30 years.

[Joe and I say] completely opposing things, but it’s the nature of the data analysis that it is very possible that we’ve come to completely opposing conclusions and they have to be worked out.

Joe is right that one of the fundamental issues is this pasture question. If you just limit yourself to cropland yield rate improvements relative to demand increases over time, it is very possible you will conclude we are not increasing fast enough to keep up. On the other hand, pasture efficiency improvements may take care of all that. There is no doubt in my mind—because I projected historical demand forward—that it probably underestimates demand. We need to work on both ends of this equation so we can come to some reasonable conclusion about what the actual trend is.”

Lead Scientist, North America Region, The Nature Conservancy

“It makes a big difference what you assume yields will do. If you look at the historic data, global demands for food are increasing faster than yields, so global cropland is increasing at around 12.4 million acres per year. At present, biofuels are expanding cropland and that land has to come from somewhere.

Part of the discrepancy is that I was looking at cropland and [John] was looking at cropland and pasture. Demand is increasing faster than yields for cropland, but that may not be the case for pasture. If cropland is expanding into pasture, the issue becomes, ‘is pasture made up by going into natural areas or not?’

One of the assumptions is how you project out yields for crops and how you project out demand. [John] projected out yields in oil crops exponentially. It’s always a problem when you project out exponentially and it doesn’t continue—you end up being really wrong as you project out further. For example, with corn yields, if you project out linearly from 1992 to 2007, you are off 1 bushel per acre; if you project out exponentially, you are off by 100 bushels per acre. You over-project if you go out too far.

The other issue is the assumptions you make about demand. Human population will go from about 6 billion to 9 billion by 2050. So a 50 percent increase. But people are eating more meat, which requires more land, so food demand will approximately double. One of the big impacts … is what happens with meat demand in developing countries? [Some projections indicate food demand may triple.]

The fact that we have controversy and debate over what future yields might be isn’t a reason not to address the issue.”

MARK NEUZIL is a professor in the Department of Communication and Journalism at the University of St. Thomas, St. Paul. He is a regular contributor to MinnPost.com and the author of five books with environmental themes.