Shifting pathways of tropical land use and their implications for carbon emissions

Tropical deforestation is increasingly recognized as one of the greatest global environmental changes induced by humans over the last half century. Forest clearing is responsible for roughly a quarter of worldwide anthropogenic carbon emissions and is considered the primary threat to global biodiversity. However, despite growing scientific attention, substantial uncertainties in the rates and patterns of deforestation and in the amount of carbon released from clearing remain.This dissertation brings together key components of tropical land use science to address major uncertainties in the global carbon budget and prospects for future deforestation and agricultural expansion. I developed improved forest biomass maps to estimate forest carbon storage and emissions with increased accuracy. I also analyzed the comprehensive Landsat database created by the United Nations Food and Agricultural Organization to provide the first detailed quantification of land use following deforestation, as well as the sources for newly expanding agricultural land. Forest conversion to cropping systems was evident throughout the tropics with more than 60 percent of cleared land entering permanent, large-scale agriculture. Indeed, the amount of forest converted to agriculture increased by 20 percent between the 1980s and 1990s. More than half of newly expanding croplands came from forests.The trends documented here indicate that the causes of deforestation have shifted from more subsistence-driven agriculture to larger-scale enterprise-driven agriculture, particularly in South America and Southeast Asia. These increasingly globalized drivers forest clearing are being accelerated, at least in part, by demand for biofuels such as ethanol and biodiesel produced from food and feed crops. In order to address the potential impact of biofuel crop production I integrated the forest biomass maps, along with global data on cropland productivity, and the improved understanding of land use dynamics. Expansion of biofuels into highly productive ecosystems will always lead to net carbon emissions for several decades, and even centuries, while expansion into degraded will provide almost immediate carbon savings. I conclude that under no foreseeable conditions will meaningful carbon benefits be achieved if crop-based biofuels are produced at the expense of tropical forests.