Removal of corn residue for biofuels can decrease soil organic carbon (SOC; refs 1, 2) and increase CO 2 emissions because residue C in biofuels is oxidized to CO 2 at a faster rate than when added to soil. Net CO 2 emissions from residue removal are not adequately characterized in biofuel life cycle assessment (LCA; refs 6, 7, 8). Here we used a model to estimate CO 2 emissions from corn residue removal across the US Corn Belt at 580 million geospatial cells. To test the SOC model, we compared estimated daily CO 2 emissions from corn residue and soil with CO 2 emissions measured using eddy covariance, with 12% average error over nine years. The model estimated residue removal of 6 Mg per ha -1 yr -1 over five to ten years could decrease regional net SOC by an average of 0.47-0.66 Mg C ha -1 yr -1. These emissions add an average of 50-70 g CO 2 per megajoule of biofuel (range 30-90) and are insensitive to the fraction of residue removed. Unless lost C is replaced, life cycle emissions will probably exceed the US legislative mandate of 60% reduction in greenhouse gas (GHG) emissions compared with gasoline.
