Premium
Model and Sensor‐Based Recommendation Approaches for In‐Season Nitrogen Management in Corn
Author(s) -
Thompson L. J.,
Ferguson R. B.,
Kitchen N.,
Frazen D. W.,
Mamo M.,
Yang H.,
Schepers J. S.
Publication year - 2015
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj15.0116
Subject(s) - sowing , growing season , nitrogen , canopy , limiting , environmental science , agronomy , seeding , mineralization (soil science) , randomized block design , crop , fertilizer , soil science , chemistry , soil water , biology , ecology , engineering , mechanical engineering , organic chemistry
Nitrogen management for corn ( Zea mays L.) may be improved by applying a portion of N in‐season. This investigation was conducted to evaluate crop modeling (Maize‐N) and active crop canopy sensing approaches for recommending in‐season N fertilizer rates. These approaches were evaluated during 2012–2013 on 11 field sites, in Missouri, Nebraska, and North Dakota. Nitrogen management also included a no‐N treatment (check) and a non‐limiting N reference (all at planting). Nitrogen management treatments were assessed for two hybrids and at low and high seeding rates, arranged in a randomized complete block design. In 9 of 11 site‐years, the sensor‐based approach recommended lower in‐season N rates than the model (collectively 59% less N), resulting in trends of higher partial factor productivity of nitrogen (PFP N ) and higher agronomic efficiency (AE) than the model. However, yield was better protected by the model‐based approach. In some situations, canopy sensing excelled at optimizing the N rate for localized conditions. With abnormally warm and moist soil conditions for the 2012 Nebraska sites and presumed high levels of inorganic N from mineralization, N application was appropriately reduced, resulting in no yield decrease and N savings compared to the non‐limiting N reference. Depending on the site, both recommendation approaches were successful; a combination of model and sensor information may optimize in‐season decision support for N recommendation.