| description abstract | Crop production depends not only on the yield but also on the area harvested. The yield response to climate change has been widely examined, but the sensitivity of crop land use to hypothetical climate change has not been examined directly. Crop land-use regression models for estimating crop area indices (CAIs)-the percent of land used for corn, soybean, wheat, and sorghum production-are presented. Inputs to the models include available water-holding capacity of the soil, percent of land available for rain-fed agricultural production, annual precipitation, and annual temperature. The total variance of CAI explained by the models ranged from 78% for wheat to 87% for sorghum, and the root-mean-square errors ranged from 1.74% for sorghum to 4.24% for corn. The introduction of additional climatic variables to the models did not significantly improve their performance. The crop land-use models were used to predict the CAI for every crop reporting district in the United States for the current climatic condition and for possible future climate change scenarios (various combinations of temperature and precipitation changes over a range of ?3° to +6°C and ?20% to +20%, respectively). The magnitude of climatic warming suggested by GCMs (GISS and GFDL) is from 3.5° to 5.9°C for regions of the United States. For this magnitude of warming, the model suggests corn and soybean production areas may decline while wheat and sorghum production areas may expand. If the warming is accompanied by a decrease in annual precipitation from 1% to 10%, then the areas used for corn and soybean production could decrease by as much as 20% and 40%, respectively. The area for sorghum and wheat under these conditions would increase by as much as 80% and 70%, respectively; the exact amount depending strongly on the change in precipitation. In general, small changes in temperature or precipitation produced larger corresponding changes (on a percentage basis) in soybean, wheat, and sorghum area than in corn area. | |
