Ecological impact of historical land-use patterns in the Great Plains: a methodological assessment

This paper demonstrates a method for using historical county-level agricultural land-use data to drive an ecosystem model. Four case study counties from the U.S. Great Plains during the 19th and 20th centuries are used to represent different agroecosystems. The paper also examines the sensitivity of the estimates of county-level ecosystem properties when using different levels of detail in the land-use histories. Using weighted averages of multiple-model runs for grassland, dryland cropping, and irrigated cropping improved prediction over a simple, single-run approach that models the prevailing land use. Model runs with the same land use and environment generally reach similar levels of soil carbon and nitrogen mineralization after ~50 years, no matter when they began, with faster convergence for irrigated cropland. Model results show that cultivation of grasslands results in large losses of soil carbon and an increase in soil nitrogen mineralization for the first 20-30 years of cultivation, which is followed by low soil carbon loss and nitrogen mineralization 50 years after cultivation started. The recently observed increase in irrigated agriculture in the central and northern Great Plains (2.7 million ha) has resulted in a net carbon storage of 21.3 Tg carbon, while irrigated cotton production has resulted in a net loss of 12.1 Tg carbon.