I read with interest the Last Word column in the October edition, written by Professor Wally Tyner from Purdue. Professor Tyner and I both appeared at the American Coalition for Ethanol (ACE) conference in Milwaukee in August and enjoyed a lively debate on the land use topic.

Naturally, I have a lot of respect for Professor Tyner and many other people at Purdue who created one of the few tools in the world (the Global Trade and Analysis Project model, or “GTAP”) that has the potential of evaluating land use changes in response to crop changes. Professor Tyner’s views can be summed by the statements:

“There is little doubt that some land use change occurs. You cannot take one-third of the U.S. corn crop and argue that there will be no land use change induced by that shift. The question then is one of measurement.”

This sounds compelling and I agree with Professor Tyner that land use changes need to be measured. And I, like Professor Tyner and many other researchers (U.S. EPA, CARB, UC-Berkeley, Iowa State University, Texas A&M, Argonne National Labs, Oak Ridge National Labs, just to name a few, not to mention all of the interested stakeholders), want to know what these land use changes are, and I believe most researchers generally are interested in making sure that how they are measured reflects the best methods and data available. But I believe the jury is still out on two issues: (1) whether there are significant land use changes induced by corn ethanol, and (2) whether the land use changes lead to significant greenhouse gas emissions. Let me explain.

First, a critical area that we must be clear about is the percent of corn land used for ethanol. Are we talking about the “gross” amount of land or are we talking about the “net” amount of land used to produce corn ethanol?

When making ethanol using a typical dry mill plant, we have livestock feed going in (corn), and ethanol and feed (distillers grains) coming out. A bushel of corn weighs 56 pounds, and yields 2.75 gallons of ethanol and 17.5 pounds of distillers dried grains from that bushel with today’s ethanol plant yields. Thus, if we assume that the 17.5 pounds of dried distillers grains replaces only corn, then the distiller grains coming out of the plant as a co-product should reduce the land use impact of the corn going into the ethanol plant by about one-third. For this reason, I believe we must always discuss the “net” amount of land rather than the “gross” amount of land needed for corn ethanol.

I believe that when Professor Tyner indicated that “you cannot take one-third of the U.S. corn crop”, he may have been referring to the “gross” and not the “net”, because the 2009 corn crop was 13 billion bushels, and 4.2 billion bushels (32 percent) will be used for producing ethanol. If that is the case, then if we correct for one-third of the corn being replaced by distillers grains, the net percent being used for ethanol is 20 percent, or one fifth, instead of one-third. The difference in acreage is between “gross” and “net” in this case is 9 million acres. The GTAP model does take this factor into account by replacing corn with distiller grains.

However, recent research on the replacement of livestock feed rations shows that distillers grains, because of their higher protein and fat content than corn, are not merely replacing corn, but are also replacing a significant amount of soybean meal in dairy cattle, poultry, and swine diets. This research shows that 1.0 pound of distillers grains is replacing almost 1.3 pounds of base feed, with about 24 percent of the base feed being soybean meal, averaged over all livestock types. Since soybean yields are much lower per acre than corn yields, the effective land use credit of distillers grains is as high as two-thirds, instead of the one-third value based on only substitution of distillers grains for corn.

Using this higher land use credit for distillers grains reduces the impact of ethanol on overall corn land use in the U.S. from Professor Tyner’s gross level of one-third to a net level of about one-tenth. The difference between “gross” and “net” land use for ethanol in this case is 19 million acres. This recent research on the use of distillers grains is not yet reflected in the current Purdue GTAP model, nor is it yet reflected in EPA’s recent land use estimates for the Renewable Fuels Standard. EPA is currently evaluating the new distillers grains research.

Moreover, there are other concerns with the GTAP model that are currently being evaluated. For example, most of the greenhouse gas emissions predicted from the model come from conversion of a relatively small amount of commercial forest (not rainforest) to cropland, as opposed to conversion of pastureland to cropland. However, the model itself excludes 61 million acres of land classified by the USDA as “cropland pasture,” which could be used for crops. If this land were included, the model would predict less forest converted and more pastureland converted, thus resulting in a lower total greenhouse gas emissions impact. GTAP researchers have acknowledged this omission and are working to incorporate this land into their model.

These are just two examples of many of the concerns with current land use estimates for corn ethanol. My point is this – the economic models, and not ground truth data, are being used to develop the various land use change estimates for corn ethanol. The land use methodology used in GTAP has only recently been developed, is still under development, and needs to be critically reviewed and tested in all areas before the predictions of land use changes and greenhouse gas emission impacts of corn ethanol can be relied upon by researchers and regulatory agencies.

 

About the Author:  Tom Darlington is President of Air Improvement Resource (AIR), Inc.