Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/23/2007
Publication Date: 5/7/2008
Citation: Solari, F., Shanahan, J.F., Ferguson, R., Schepers, J.S., Gitelson, A. 2008. Active Sensor Reflectance Measurements of Corn Nitrogen Status and Yield Potential. Agronomy Journal. 100:571-579 Interpretive Summary: Current nitrogen management strategies for cereal production systems in the U.S. and around the world are characterized by low nitrogen use efficiency, environmental contamination, and considerable ongoing public debate regarding use of nitrogen fertilizers in crop production. To address these concerns, some have advocated use of active sensor reflectance measurements of in-season corn nitrogen status for directing spatially-variable fertilizer applications. However, first it is necessary to confirm that active sensors can reliably assess corn nitrogen status. The active sensor used in our work was the Crop Circle model ACS-210 manufactured by Holland Scientific of Lincoln, NE. The sensor generates its own source of modulated light (pulsed at ~ 40,000 Hertz) in two wavebands (amber color and near infrared) using a single bank of light emitting diodes (LED), and measures the percent of modulated light reflected back from the crop canopy, using another bank of LED. The specific objectives for this research were to determine the most appropriate: 1) growth stages and 2) sensor-computed vegetation indices with the greatest sensitivity in assessing corn canopy nitrogen status along with grain yield potential. Field plots were established by supplying varying amounts of fertilizer nitrogen at crop planting. Active sensor measurements of canopy nitrogen status were then taken at four different crop growth stages (2 pre-tassel and 2 post-tassel dates). The sensor was mounted on an adjustable height boom on a high clearance vehicle at 1 meter above the canopy, interfaced to a computer, and readings were collected as the vehicle traveled through the plots. An independent assessment of leaf N status was also taken at the same time sensor readings were acquired, using the hand-held Minolta SPAD chlorophyll meter. Sensor readings were converted to the three different vegetation indices 1) amber normalized difference vegetation index (ANDVI), 2) simple ratio (SR) and 3) chlorophyll index (CHLI), using three different mathematical formulae involving the sensor amber and NIR reflectance values to compute the three indices. Grain yields were also measured. Results from this study showed that the sensor-computed vegetation indices were more highly correlated with SPAD chlorophyll meter readings for the two pre-tassel stages than the two post-tassel stages, which was attributed to the inability of sensor readings to detect canopy nitrogen status due to interference from tassels on the plants during reproductive growth. Of the three sensor-computed vegetation indices (ANDVI, SR, CHLI), SR and CHLI values were found to be more sensitive than ANDVI in separating differences in canopy nitrogen status and grain yield potential. Based on these results, we concluded that sensor readings acquired during vegetative growth and expressed as either the SR or the CHLI would have the greatest potential for differentiating canopy nitrogen status and directing spatially-variable in-season nitrogen applications. However, first it will be necessary to validate our results in a wider range of soils, climate and geographical conditions, and develop algorithms for translating sensor reading into appropriate nitrogen fertilizer application rates.
Technical Abstract: Use of active crop canopy sensor reflectance measurements of in-season corn (Zea mays L.) nitrogen (N) status for directing spatially-variable N applications has been advocated to improve N use efficiency. However, first it is necessary to confirm that active sensors can reliably assess N status. Our research goals were to determine the most appropriate: 1) growth stage and 2) vegetation index with greatest sensitivity in assessing canopy N status and grain yield, using active sensor readings. Variable crop N conditions were generated by supplying fertilizer N at different amounts and times in three field studies conducted near Shelton, NE in 2005. Sensor and SPAD chlorophyll meter readings were gathered at two vegetative (V11 and V15) and two reproductive (R1 and R3) growth stages, using the Crop Circle active sensor that measures canopy reflectance in two bands (amber and NIR; centered at 590 and 880 nm). Reflectance values were converted to three vegetation indices; the amber normalized difference vegetation index (ANDVI), simple ratio (SR), and chlorophyll index (CHLI). Grain yields were also determined. Variation among N treatments for all vegetation indices was more highly correlated with SPAD readings for vegetative than reproductive growth stages, with the SR and CHLI being more sensitive than ANDVI in detecting variation in canopy greenness. The SR and CHLI values were also more sensitive in distinguishing grain yield variation than ANDVI. Our findings indicate active sensor assessments are capable of detecting variations in canopy N status and could be used to direct spatially variable N applications.