Land conversion and greenhouse gas emissions
Reducing greenhouse gas emissions in food, land, and agriculture (FLAG) sectors requires reducing the emissions from crop production, which are largely connected to fertilizer use, power for irrigation, and management operations. In addition, converting grasslands, forests, or wetlands to row crop agriculture generates a significant level of emissions. Companies may also be asked to make "no deforestation" or "no land conversion" commitments. Because land conversion and the associated emissions occur in the supply chain outside a corporation’s direct control, they are difficult to measure precisely, particularly in commodity crop supply chains, which are rarely directly traceable.
FoodS3 provides a way to estimate exposure to land conversion and calculate the associated emissions for direct land use change in the contintental United States - referred to as land use change (LUC) emissions. We use county-level production, land use, and crop practices data to calculate emissions for commodity crops, and use an optimization model to estimate the upstream sourcing for a given facility, company, or sector.
FoodS3 focuses on corn, soy, and wheat, the top three crops produced in the US by area. These crops are mostly used as feed for animals, for biofuels like ethanol, or exported. Emissions from the production of feed and feedstocks make up a majority of the emissions embedded in broiler chicken, fed beef cattle, hog, and ethanol production, but land conversion contributes differently to each industry. For broiler chickens, 44% of the emissions from feed crops come from land use change. For ethanol, by contrast, the percentage is only 16%. Fed beef (18%), and hog (22%) production are between those two extremes.
These differences come from two main sources: the mix of crops used as feed or feedstock, and the locations supplying those crops.
If we look at the average emissions resulting from the production of one bushel of corn, soy, and wheat, we can see that they differ both in the total amount of greenhouse gasses released as well as the fraction that comes from LUC.
Corn has by far the highest per-bushel emissions, 80% of which come from production. Corn requires high fertilizer rates and often uses irrigation (which usually requires energy - either electricity or diesel - for pumping), both of which lead to GHG emissions. By contrast, soybean production uses much lower rates of nitrogen fertilizer, since soybean, a legume, can fix nitrogen from the air. The land use change emissions for corn and soy are similar in absolute terms, but because soybean production emissions are much lower, LUC emissions make up over half the emissions from soybean.
Broilers, because they are fed a high-protein diet, use more soy in diet rations than other livestock, which contributes to the greater impact of LUC emissions in that industry. Ethanol, by contrast, uses only corn, which has the lowest fraction of emissions from land use change.
Differences in emissions between crops explain part of the difference between broilers and ethanol, but even if we look at just corn, not all corn is created equal. If we look at only the emissions from corn, we still see differences among industries.
The corn going to broiler chickens has higher emissions per bushel, and a larger percentage of LUC emissions, than the corn that goes to ethanol. This is mainly because ethanol plants and broiler facilities source from different locations.
If we look across the USA, some areas have more cropland expansion than others - in this case, we are looking at the acres converted to corn.
Land use change rates are low in the traditional Corn Belt region, where most suitable land is already used for corn and soy. However, we can see increased land use change in the northern and western Great Plains region, and in the east, particularly in Appalachia.
Based on the flow estimates from FoodS3 we can see that corn for ethanol is mostly grown in the Corn Belt. This is also the area where yields are highest, so on a per-bushel basis, emissions remain relatively low.
By contrast, the corn fed to broilers is sourced quite widely, with a larger fraction coming from the Southeastern US, which has a higher concentration of broiler facilities. These areas have higher rates of land use change, while yields are lower, contributing to higher overall emissions per bushel.
A note on attribution
In the current version of the model, LUC is attributed to the breakout crop - the first crop grown on newly cleared land. Future versions will use a weighted average for attributing LUC to each crop grown in the rotation. Our initial estimates show similar results using this new method.
Attribution for corn fed directly to animals is straightforward. For ethanol production, we use an economic allocation between ethanol and DDGS, where the DDGS fraction is attributed to consumers of that DDGS, usually livestock and poultry. Soybeans pass through crushing facilities, where again an economic allocation is used to attribute a percentage of the emissions and LUC to soy oil and to soy meal. The emissions for livestock consuming soy meal contain only that fraction allocated to soy meal.
This brief was created in partnership with The Nature Conservancy. The publication of this material was made possible through funding by the Walmart Foundation. The findings, conclusions, and recommendations presented in this brief are those of FoodS3 alone and do not necessarily reflect the opinions of the Walmart Foundation.