Studies on factors affecting the evolution of agroecosystems in the Dakotas
Peter T. Wolter
This dissertation combines remote sensing and applied economics tools to study land use conversions in North Dakota and South Dakota that are tied to this region’s overall socio-economic welfare. Specifically, the region’s corn and soybeans cultivation expanded significantly over the past decade replacing the region’s grasslands and grain crops. In paper I, we estimate the localized impacts of the advent of corn-based ethanol plants on the Dakotas’ corn acreage. We implement a Difference-in-Difference framework through more flexible assumptions as the Parallel Paths assumption of the standard model fails to hold. We find strong trends in the Dakotas’ corn acreage over the past decade, but surprisingly some ethanol plants were found to have a negative impact on local corn acreage. In paper II, we evaluate crop competitiveness due to heterogeneous weather impacts on crop yields, and then test whether annual weather fluctuations explain land allocations among the Dakotas’ major land uses. Our integrated framework suggests that annual weather variability is an important determinant of regional land use decisions. Under the A1B emissions scenario of climate change, we find that the yields of all of the Dakotas’ major crops will decline by 2031-2060 relative to 1981-2010, leading to lower (higher) spring wheat (alfalfa) acres in Eastern (Western) Dakotas. In paper III, we develop and implement a satellite image-processing algorithm to estimate historical land use acres using raw Landsat sensor data, thereby extending the existing Cropland Data Layers back to 1984 in eastern Dakotas. We demonstrate that the availability of a longer time-series is useful as the rate of land use change may differ among different time-spans. In paper IV, we evaluate the cost-effectiveness of grassland conservation easements when spatial spillovers are present among private landowners. We first develop a conceptual model to incorporate social spillovers in evaluating the role of easements in inhibiting grassland conversions. We empirically test whether social spillovers are present by estimating hazard rates of conversion as a function of neighborhood density of grasslands and easements. Our findings suggest that easements are strategic complements to existing grasslands in preventing grassland conversions in the Dakotas.