Sensor‐Based Nitrogen Applications Out‐Performed Producer‐Chosen Rates for Corn in On‐Farm Demonstrations

Published in Agron. J. 103:1683–1691 (2011) Posted online 15 Sep 2011 doi:10.2134/agronj2011.0164 Copyright © 2011 by the American Society of Agronomy, 5585 Guilford Road, Madison, WI 53711. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. E optimal nitrogen fertilizer rate (EONR) for corn and other crops can vary substantially within fields (Schmidt et al., 2002; Mamo et al., 2003; Scharf et al., 2005; Kitchen et al., 2010) and among fields (Schmitt and Randall, 1994; Bundy and Andraski, 1995). Current N management practices do not address this variability. Most U.S. corn producers apply the same rate of N fertilizer to whole fields and often to whole farms. The adoption of tools to diagnose fertilizer N need has been slow (Kitchen and Goulding, 2001). In the past 10 yr, prices for both N fertilizer and corn have increased substantially, increasing the financial incentive to apply no more N than the crop needs but also to make sure that N is supplied in sufficient amounts. In addition to the agronomic and economic benefits, diagnosing and applying the EONR produces environmental benefits by reducing nitrate levels left in the soil after harvest (Hong et al., 2007). Accurate, convenient, and affordable methods to diagnose EONR are needed now more than in the past. Understanding why the EONR varies will help us to devise more effective strategies for managing N. It appears that variability in crop yield and demand for N is usually not a major factor determining the EONR (Lory and Scharf, 2003; Nafziger et al., 2004; Scharf et al., 2006b). This leaves variability in the soil N supply as the probable controlling factor, although it would be desirable to have a body of evidence that directly supports this hypothesis. Many lab tests for soil N availability have been devised, and some have performed well in small data sets, but in large data sets they have performed poorly (Scharf et al., 2006a; Laboski et al., 2008). It has long been known that N-deficient corn reflects more visible and often less near-infrared light than N-sufficient corn (Walburg et al., 1982). Scharf et al. (2006a) showed in a large and geographically-dispersed study that chlorophyll meter (transmittance) measurements of corn leaves provided much better prediction of the EONR than any of 26 soil N tests. Improved diagnostic accuracy is the main justification for pursuing canopy-sensor-based N management in preference to soil-test-based management. Technological advances have enabled us to use spectral measurements of crops to diagnose and control N fertilizer rates. Reflectance sensors have logistical advantages over other potential spectral measurements. They can manage spatial variability in N need more easily than hand-held meters, and can operate under conditions that prevent the acquisition of aerial images. These advantages have led to the development of methods to translate sensor data into N rate decisions (Mullen et al., 2003; Dellinger et al., 2008; Scharf and Lory, 2009; Barker and Sawyer, 2010; Kitchen et al., 2010). Although there is still a need for decision systems to be improved and differences between them resolved, reflectance sensors are commercially available to guide variable-rate N applications. They can diagnose crop N needs and control N application rates at a fine spatial scale. The expected benefits are identification of places where N rate can be reduced without hurting yield; the identification of places where ABSTRACT Optimal N fertilizer rate for corn (Zea mays L.) and other crops can vary substantially within and among fields. Current N management practices do not address this variability. Crop reflectance sensors offer the potential to diagnose crop N need and control N application rates at a fine spatial scale. Our objective was to evaluate the performance of sensor-based variable-rate N applications to corn, relative to constant N rates chosen by the producer. Fifty-five replicated on-farm demonstrations were conducted from 2004 to 2008. Sensors were installed on the producer’s N application equipment and used to direct variable-rate sidedress N applications to corn at growth stages ranging from V6 to V16. A fixed N rate chosen by the cooperating producer was also applied. Relative to the producer’s N rate, sensors increased partial profit by $42 ha–1 (P = 0.0007) and yield by 110 kg ha–1 (P = 0.18) while reducing N use by 16 kg N ha–1 (P = 0.015). This represents a reduction of approximately 25% in the amount of N applied beyond what was removed in the grain, thus reducing unused N that can move to water or air. Our results confirm that sensors can choose N rates for corn that perform better than rates chosen by producers.