German Mandrini

German is a Data Scientist at FBN, working at the intersection of data science, agronomy, and finance. He comes from a family farm, and agriculture has guided much of his career. He holds a B.S. in agronomy, an M.S in agricultural economics, and a Ph.D. in crop sciences. He has more than ten years of experience working for different companies in the farming sector, including grain trading, input selling, farm management, and corn breeding.

09 Mar 2022

by German Mandrini

The Economic Optimum N Rate (EONR) is the N rate that can help maximize profit potential for a given field. Many methods promise farmers that they would predict the EONR of their fields with higher accuracy and, consequently, increase their profits. Some are simple, require few inputs, and provide recommendations that do not change for particular years or fields and instead are designed to work well on average across a wide range of growing conditions. Among these is the Maximum Return to Nitrogen tool (MRTN), the current recommendation system promoted in the US Midwest, based on decades of real-world trials conducted by several universities in the area. Some other methods are complex and usually require specific inputs at the field level or sometimes for different soils inside the field. Every year, they provide a particular N recommendation, varying according to the soil and crop conditions. A fundamental assumption of these recommendation methods is that an improved prediction accuracy – will save N inputs when possible, improving economic and environmental outcomes. Are those complex tools better? So far, studies have shown that complex methodologies struggle to increase farmers' profits, which explains why adoption remains low. A recent study covering 49 sites and three years of trials help explain the reasons. In that study, they compared many N recommendations tools available nowadays, including tools that require site-specific soil information  (e.g., PPNT, PSNT) or complex simulations programs (e.g., Maize-N) and simpler tools that provide a stable recommendation for a broad region (e.g., MRTN). Among 31 tools, the MRTN was the tool that best maximized profitability. Additionally, they also measured the environmental outcome of the tools, shown by how much N is lost from the soil-crop system, and complex tools did not provide consistently better results in that aspect either.  Another study compared MRTN with a complex tool that used machine learning, soil sampling, weather, and crop information to improve the accuracy of the predictions. They found that the complex tool slightly increased the accuracy of the recommendation, but that did not translate into higher profits. The reason is that predicting EONR is challenging since it depends on the balance between what the crop needs and what the soil will provide, and both are unknown at the time of making a decision. The EONR is highly variable across fields, and it changes every year even in the same field. In that context, complex methods can not achieve an accuracy high enough to meet their promise of increasing profits and many times recommend N rates that are far from what the real EONR need was. Simpler methods, with static N recommendations that work well in most situations, escape that problem, and never recommend N rates far from the target, as it happens with the complex tools. From an environmental point of view, they concluded MRTN used at the low end of their recommended N range achieves similar N losses than the complex tool. So, how does MRTN work? Experts from the universities' extension service divided the area into regions with similar soil and crop growing conditions. They conducted trials with several N rates in those regions for multiple years. In each of those trials, we can obtain what is known as the profits response curve to N, which shows what the profits for increasing N rates are. Then they averaged all the profits curves for multiple trials and finally selected the N rate that maximizes the return to N -that's where the name comes from. Every year, new trials are conducted, and the MRTN calculation is updated to keep it current with new hybrids and any change in weather patterns that affect the response.  The MRTN is an improvement over old methodologies known as yield-based approaches -i.e., the 1.2 lb. N per bushel of expected yield. Those methods were developed in the 1970s and suggested that higher yield potentials require higher N rates. Over time, researchers understood that in the US Midwest, the EONR was not very much related to yield at the EONR rate, meaning that high-yielding corn needs more N but not necessarily more Nitrogen fertilizer. Part of the reason is that conditions that lead to higher yields, like higher rain and temperature, also lead to higher mineralization which increases the N provided by the soil and reduces the need for fertilizer. Seeing this lack of relationship, researchers decided to move in a different direction, and that's how they created the MRTN, which is based on the N rate that maximized profits in a region instead of focusing on yield.  In summary, the MRTN is a methodology based on real data that has proved to maximize profits, even when compared with tools that use advanced technology to provide site-specific recommendations. The MRTN is also an improvement over yield-based methods that tend to recommend higher N rates than actually needed, reducing profits for farmers. *Mineralization: The release of nitrogen from soil organic matter in a form that plants can use. Copyright © 2014 - 2022 Farmer's Business Network, Inc. All rights Reserved. The sprout logo, Farmers First flag logo, "Farmers Business Network," "FBN," and "Farmers First" are registered trademarks of Farmer's Business Network, Inc. or its affiliates. All other trademarks are the property of their respective owners. Terms and conditions apply.