Land-Use Changes and CO2 Emissions Due to US Corn Ethanol Production due to higher prices Hertel et al (2010) provide decomposed contributions of these factors in mitigating the land-use impacts of ethanol production The land required to increase the US ethanol production from its 2001 level to 15 BGs obtained from these simulations is smaller than its corresponding value reported in Tyner et al (2009) by about 16.7% (i.e., 2.96 vs 3.55 million ha) Two major modifications in the GTAP model contribute to this reduction A portion of this reduction is associated with the land conversion factors As noted earlier, a new set of regional land conversion factors at the AEZ level is used in this work The new land conversion factors in several AEZs are higher than the single conversion factor of 0.66, which was used in earlier work (for details see Tyner et al., 2010) Introducing the new land categories (cropland pasture and unused land) (in these simulations, the area of US CRP land is held constant) into the model also contributes to the reduction in land requirement In particular, in the USA and Brazil, in the presence of cropland pasture, farmers convert a portion of this type of land to crop production For example, an increase in US ethanol production from its 2001 level to 15 BGs leads to the conversion of 1.2 million from cropland pasture to cropland Indeed, a portion of this land conversion prevents sharp reductions in production of other crops It is important to note that the competition between crop and livestock industry prevents full conversion of cropland pasture to crop production These two modifications not only reduce the land requirement of ethanol production, but also alleviate the adverse impact of ethanol production on the prices and consumption of crops Table also indicates that the required land for producing 1000 gallons of ethanol grows with higher levels of ethanol production For example, for the 2001–2006 simulation, an additional 3.085 BGs of ethanol triggers global land-use changes of roughly 610 This is equal to 0.20 per 1000 gallons of ethanol However, for the 13–15 BGs simulation, an additional 1000 gallons of ethanol requires 0.25 of land To increase ethanol production from the 2001 level to 15 BGs, on average about 0.22 of land per 1000 gallons of ethanol are required The marginal level (0.25) is higher than the average (0.22), which would be expected because as more land comes into production, the yields on the incremental area would be lower Table depicts another aspect of the land-use implications of US ethanol production This table shows the distribution of Table a land-use changes between forest and grassland About 25% of the required croplands needed to increase ethanol production from its 2001 level to 15 BGs come from forest, and the remaining (75%) come from grasslands Table also indicates that with higher levels of ethanol production, the portion of forests in the converted land into crop production increases very slightly (from 23% in 2001 to 25% at the 15 BGs ethanol production) In the absence of crop yield growth, the increasing global land-use change, given equal increments of US ethanol production, is explained by the differences in the productivity of available lands Productive lands are employed first before marginal lands, which have lower productivity and lower yields At low levels of production, more productive lands are available; hence, less land is required to produce additional ethanol However, at higher levels of ethanol production, most of the productive land is already being used, and only marginal land is available Given this, more marginal land is required to produce the same increment of US corn ethanol production Group 2: Simulations with Updated Baseline for the Time Period 2001–2006 The global economy changed significantly over the 2001–2006 period Countries followed different economic growth paths, population increased everywhere at different rates, land productivity rapidly increased in many regions (with some exceptions), and technology improved in many areas These are important factors that could alter the land-use implications of biofuels In the second group of simulations, these factors are taken into account To accomplish this task, a database was developed to include data on crop production, harvested area, forest areas, gross capital formation, labor force (skilled and unskilled), gross domestic product, and population for the whole world at the country level Table contains the percentage changes in macroeconomic variables (for details see Tyner et al (2010)) Then this data set was used to generate a baseline that replicates the transition of the global economy from 2001 to 2006, whereas global biofuel production during this time period was targeted in the presence of population, income, and yield growths In building the baseline, the model was guided to replicate the historical paths of changes in harvested Global land-use changes due to US ethanol production: off the 2001 database Changes in US corn ethanol output 3.085 BG 2.145 BG 2.000 BG 2.000 BG 2.000 BG 2.000 BG 13.23 BG 545 (2001–2006) (2006–7 BG) (7–9 BG) (9–11 BG) (11–13 BG) (13–15 BG) (2001–15 BG) Land-use changes (ha) Distribution of land-use changes (%) a Forest Grassland Crop À 143,716 À 114,409 À 112,330 À 116,795 À 120,688 À 124,151 À 732,089 À 466,652 À 345,912 À 335,712 À 347,864 À 359,650 À 371,156 À 2,226,946 610,376 460,324 448,041 464,657 480,345 495,311 2,959,053 Forest Grassland Totala 23.5 24.9 25.1 25.1 25.1 25.1 24.7 76.5 75.1 74.9 74.9 74.9 74.9 75.3 100.0 100.0 100.0 100.0 100.0 100.0 100.0 The difference between the changes in cropland and the sum of forest and grassland is due to rounding Cropland pasture is included in cropland