Agriculture in a Changing Landscape: Modeling shifts in the geospatial distribution of crops in response to climate change

Abstract

Altered patterns of temperature and precipitation associated with global climate change are expected to affect the productivity of agricultural regions around the world, with varying regional impacts. Since ideal environmental conditions vary depending on the physiological needs of specific plant types, the regions where we grow different crop varieties are likely to shift in response. This shift will have profound implications for rural landscapes and communities, as well as global food supply and international markets. In this research I use Classification and Regression Tree (CART) modeling to investigate whether changes in climate over the past 50 years have contributed to shifts in the distribution of crops in Minnesota. I incorporate climate, soil, and agricultural management data to create a time series of regression tree models which predict the acreage of three different important commodity crops, corn, soy, and wheat, for each county. The resulting models indicate that farmers’ decisions to grow corn are positively associated with warmer winter temperatures, and that the temperature threshold has increased over time. Soil quality is the primary predictor of soybean acreage, with a stable threshold over time. Wheat models produced inconsistent results, possibly due to displacement by conversion of wheat acreage to corn acreage. This suggests that farmers are already employing crop-switching strategies in response to recent changes in climate. As the impacts of climate change increase in severity, additional research and investment will be needed to help agricultural producers continue to adapt.