Herein, some of the many specific factors that affect scenarios of land-use emissions have been discussed. From a correlation analysis that compared the influence of changing population, economic activity, and technological change on land-use emission scenarios, Alcamo and Swart (1998) concluded that population was the most influential driving force. The reason is the relationship between population and increasing food demand, which leads to more cows that produce CH4 and more extensive fertilized croplands that release N2O. Although most scenarios also assume that improvements in crop and animal productivity will partly compensate for increasing food demand, some authors do not believe that productivity increases can be sustained. For example, Kendall and Pimentel (1994) show a decrease, rather than increase, in per capita grain production because of less optimistic assumptions about the increase in crop productivity (0.7% per year). Brown and Kane (1995) point out some developments that may slow down productivity:
Brown and Kane (1995) argue that in recent years rice yields have either stabilized or fallen in many key rice-producing countries, and suggest that dramatically boosting rice yields above a level of 4 tons per hectare may require new technological advances. If the skeptics are right, and the assumed productivity increases of the "business-as-usual" scenarios are not realized, then a greater expanse of cropland would be needed to satisfy the same agricultural demand. This expansion would lead to still higher emissions of CH4 from rice fields and livestock, and more rapid deforestation and earlier peaks in the emissions of CO2 . However, the lower increase in productivity probably also implies lower fertilizer use per hectare, which may lower emissions of N2O from fertilized soils.
Other reports in this collection