Skip to main content
ARS Home » Midwest Area » Columbia, Missouri » Cropping Systems and Water Quality Research » Research » Publications at this Location » Publication #377167
Research Project: Sustainable Intensification of Cropping Systems on Spatially Variable Landscapes and Soils

Location: Cropping Systems and Water Quality Research

Title: Improving publicly available corn nitrogen rate recommendation tools with soil and weather measurements

Author
item Ransom, Curtis
item KITCHEN, NEWELL
item SAWYER, JOHN - IOWA STATE UNIVERSITY
item CAMBERATO, JAMES - PURDUE UNIVERSITY
item CARTER, PAUL - FARMER
item FERGUSON, RICHARD - UNIVERSITY OF NEBRASKA
item FERNANDEZ, FABIAN - UNIVERSITY OF MINNESOTA
item FRANZEN, DAVID - NORTH DAKOTA STATE UNIVERSITY
item LABOSKI, CARRIE - UNIVERSITY OF WISCONSIN
item MYERS, BRENTON - CORTEVA AGRISCIENCE
item NAFZIGER, EMERSON - UNIVERSITY OF ILLINOIS
item SHANAHAN, JOHN - SOIL HEALTH INSTITUTE

Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/25/2021
Publication Date: 2/8/2021
Citation: Ransom, C.J., Kitchen, N.R., Sawyer, J.E., Camberato, J.J., Carter, P.R., Ferguson, R.B., Fernandez, F.G., Franzen, D.W., Laboski, C.A., Myers, B.D., Nafziger, E.D., Shanahan, J.F. 2021. Improving publicly available corn nitrogen rate recommendation tools with soil and weather measurements. Agronomy Journal. 113(2):2068-2090. https://doi.org/10.1002/agj2.20627.
DOI: https://doi.org/10.1002/agj2.20627

Interpretive Summary: Applying nitrogen (N) fertilizer at the rate sufficient for crop N needs, but not more, can improve farmers’ profits and help reduce loss of N off agricultural fields. Previous research shows that current N fertilizer recommendation tools for corn are not adaptive to variable soil and weather. This research looked at improving 21 N rate recommendations tools using research conducted on 49 fields across eight U.S. Midwest states. Using two machine learning algorithms— previously tested for their performance and interpretability when adjusting tools—we tested each tool for improvement by incorporating site-specific soil and weather measurements. All tools were improved with at least one of the two machine learning algorithms, albeit, some tools had minimal improvements. Incorporating the evenness of rainfall, soil pH, soil carbon, and bulk density were most helpful for improving tools. Tool improvement reduced the over-application of N fertilizer, thereby minimizing N related environmental degradation. Additional research is required to ensure these adjustments are robust year after year. Farmers will benefit from improved N recommendation tools because both over- and under-applying N is uneconomical. Also, when over-applications of N fertilizer are reduced, N loss to lakes and streams will be reduced.

Technical Abstract: Improving corn (Zea mays L.) N fertilizer rate recommendation tools is necessary for improving farmers’ profits and minimizing N pollution. Research has repeatedly shown that weather and soil factors influence available N and crop N need. Adjusting available corn N recommendation tools with soil and weather measurements could improve farmers’ ability to manage N. The aim of this research was to improve publically available N recommendation tools with site-specific soil and weather measurements. Information from 49-site years of N response trials in the U.S. Midwest was used to evaluate 21 rate recommendation tools for single (at-planting) and split (at-planting + sidedress) N applications. Using elastic net and decision tree algorithms, the difference between each tool’s N recommendation and the economically optimum N rate (EONR) was modeled against soil and weather measurements. The model’s predicted values were used to adjust the tools. Unadjusted, the best performing tool had r^2 = 0.24; after adjustment the best performing tool had r^2 = 0.57. Overall tool improvement was modest, at best, and sometimes required many additional inputs. Using weather measurements (e.g., evenness or rainfall or abundant and well-distributed rainfall) helped increase N recommendations by accounting for N loss while soil measurements (e.g., pH and total C) helped decrease N recommendations when there was sufficient available soil N. This investigation showed that incorporating soil and weather measurements is a viable approach for improving corn N recommendation tools on a regional basis; but even with adjustments, tools still had room for additional improvement.