Observed climate variability and change impacts on agricultural productivity in the Midwestern US

Agricultural production has thus far succeeded in keeping pace with the demands of a growing world population; however, changes in climate and climate variability, increased demand for agriculture to support the development of biofuel resources, and continued growth in the world population call into question the ability of global agricultural production to satisfy requirements. The US Cornbelt alone produces 36% and 29% of the world's corn and soybean supply, respectively, and encompasses a wide range of climates and agricultural management practices. Understanding the potential impact of climate variability and change on agricultural production is required for developing management strategies to mitigate the effects of projected future climate change. This study makes use of observations of precipitation and air temperature, a large-scale hydrology model and crop production information collected by the United States Department of Agriculture's National Agricultural Statistics Service (NASS) to evaluate the sensitivity of crop production to historic climate variability and trends between 1915 and 2009. Climate data was gridded and used to run the hydrology model to provide spatial and temporal information on parameters such as soil moisture, soil temperature, snow cover, runoff and agricultural drought severity. These variables were evaluated for statistically significant trends and correlations with annual crop yields. Crop yields were first de-trended to remove the effects of changes in genetics and fertilizer applications. Then additional metrics, such as the soil temperature and moisture at planting and harvest and the occurrence of extreme heat or precipitation during crop development, were correlated with crop yield to identify the climate conditions to which crop productivity is most sensitive. Finally, trends in these metrics and their role in controlling the variability in crop production, especially in maintaining minimum yields, were analyzed. In many cases, metrics to which crops are most sensitive actually displayed trends not visible in the analysis of more common climate variables such as annual average precipitation and air temperature.