Graduate Dissertation Defense - Paul Hegedus

Apr 27 2022 9:00 am
Zoom Virtual Meeting

Dissertation Defense Information

Paul Hegedus

April 27th 9am – 10am

Leon Johnson Hall Room 325

Title: Optimizing Site-Specific Nitrogen Fertilizer Management Based on Maximized Profit and Minimized Pollution


Application of nitrogen fertilizers beyond crop needs can contribute to nitrate water pollution and soil acidification. Excess nitrogen applications are most prevalent where synthetic fertilizers are applied at uniform rates across fields. Precision agroecology harnesses the tools and technology of variable rate precision agriculture, a common but underutilized management strategy, to make ecologically conscious decisions about field management that promote economic and environmental sustainability.

The objectives of this dissertation were to 1) outline the conceptual basis for performing on-farm precision experimentation to aid in agroecological management, 2) understand the field-specificity in crop responses to variable nitrogen fertilizer rates, 3) identify how statistical forms of crop response models vary between fields and across years, 4) assess data constraints related to crop response modeling, 5) evaluate and develop a model for nitrogen use efficiency in dryland winter-wheat agroecosystems, 6) develop optimized nitrogen fertilizer management on a subfield scale based on maximization of farmer net-returns and nitrogen use efficiency.

On-farm precision experimentation provides the basis for making ecological data driven management decisions through the field-specific assessment of crop responses. The relationship between winter-wheat yield and grain protein concentration to varying rates of nitrogen fertilizer varies between fields, and across time. The specificity in how crops respond to nitrogen fertilizer across space and time influences the method used to characterize the relationships between crop production and quality to nitrogen fertilizer in a field. Machine learning based approaches, such as random forest regression, tend to provide the best model for forecasting future crop responses. Data are available to farmers sooner and at more frequent time intervals due to the advances in sensors and technology related to precision agriculture. Using information that is collected from open-source data until a farmer needs to make decisions on management does not always improve the quality of crop predictions, however never reduced the accuracy of predictions. Patterns between nitrogen use efficiency and nitrogen fertilizer are similar between dryland winter-wheat fields. Machine learning again demonstrated its capacity to be utilized in agronomic situations, as a support vector regression model provided the best predictions of nitrogen use efficiency on a subfield scale. When integrating efficiency into the decision-making framework for nitrogen fertilizer management, site-specific optimized fertilizer management, based on maximized profits and minimized potential of nitrogen loss, increased profitability compared to farmer’s status quo management 100% of the time on average across all fields tested. However, even when considering nitrogen use efficiency into the identification of optimum nitrogen fertilizer rates, improved sustainability with a site-specific approach compared to a farmer’s status quo management was field specific.

Link: https://us02web.zoom.us/j/81405360985?pwd=a0FIK1Q3UnF1U051ZXhmaFpweUVaUT09