Persistence in tillage decisions: Aggregate data analysis

The Emission Gap Report 2013 from the United Nations Environment Program showed that adopting conservation tillage such as no-till, as an alternative to conventional tillage, contributes significantly to climate change mitigation through carbon sequestration. However, substantial amounts of soil carbon are lost when farmers interrupt continuous use of conservation tillage with conventional tillage. Conservation tillage is spreading, but little is known about the behavioral persistence in tillage decisions. To address the gap in the literature, we estimate county-specific Markov models of tillage-crop choices, and use the predicted probabilities of alternative two- and three-year tillage rotations to evaluate spatial variation and temporal persistence in conservation tillage adoption for the state of Iowa (U.S). We find that the county-average probabilities of continuous conservation tillage range between 0.133 and 0.295, and vary significantly among crop rotations. We also find a statistically strong positive effect of the incidence of the highly erodible land on the county-average use of continuous conservation tillage. Our results underscore the importance of dynamic modeling for understanding behavioral persistence in tillage decisions, and the interdependence between farmers' crop and tillage rotations. Keywords: Aggregate data, Continuous conservation tillage, Markov chain model, Crop rotations, Highly erodible lan