Globally, there has been large-scale conversion of natural grassland to cropland ecosystems which this has led to land degradation that could reduce future food security, other ecosystem services and even climate. Currently, there is a dearth of quantitative information assessing the severity, distribution, and causes of this land degradation. For practical purposes, this information is needed to develop improved methods of land use (LU) conversion. Uruguay, in contrast with many other regions, still has a high proportion of unimproved grasslands but, during the last 15 years, there has been extensive conversion to grow grain crops.

The fundamental goal of this dissertation was to quantify soil degradation resulting from this LU change. Two aspects of soil degradation were studied, soil organic carbon (SOC) and erosion by water. The Environmental Policy Integrated Climate biophysical simulation model (EPIC) was used to model the grassland and cropping systems. The study consisted of three steps: (1) calibration and validation of the model for the Uruguayan agroecosystems, and development of a spatial version, (2) identification of the LU change areas, and (3) quantification of soil degradation as a result of the LU changes.

The EPIC model adequately reproduced the field-scale SOC dynamics and erosion in field validation sites. Further, the spatial version of the model was found to simulate spatial and temporal performance adequately. LU change areas during 2000-2013 were mapped and found to cover an area of 410,000 ha, about 13% of potential area for commercial agriculture. LU greatly affected soil degradation. It was greatest for continuous Soybean cultivation with no crop rotation, and lowest for grassland (no conversion to cropping). In addition to LU, slope and initial SOC had significant effects on degradation.

The main conclusions were that the recent and continuing conversion from grassland to cropland has caused significant soil degradation, but that some modifications of LU can reduce the risk of degradation.