Model for detecting nitrogen deficiency in wheat crop using spectral indices

Abstract

An experiment was conducted four years (2007-08 to 20010-11) during rabi at Anand, India to study nitrogen stress response of wheat (Triticum aestivum L.) on spectral signature using spectroradiometer (Model - UniSpec DC Dual Channel Spectrometer, PP System, USA). The experiment was carried out with two wheat cultivars viz., GW496 and Lok 1 and five levels of nitrogen 120, 90, 60, 30 and 0 kg ha-1. The NDVI (normalized difference vegetation index) which is an indicator of status of crop biomass was found higher at 45 to 65 DAS of the crop age. The same characteristics were also observed in case of NDRE (normalize difference red edge) index at the 45 to 65 DAS of the crop age. Based on the index value of NDVI & NDRE at different growth stages of wheat crop, it was revealed that the crop biomass under the 120 N kg ha-1 could be more healthy which resulted in better yield. The variety wise analysis showed that no significant varietal differences were observed for NDVI, mSR and NDRE during different growth stages of wheat crop. The index value of NDVI and NDRE was higher in the highest nitrogen 120 kg ha-1 and lowest in treatment without nitrogen application. The NDVI was found more related to biomass production, while the NDRE was found useful to detect nitrogen stress during 45 to 75 DAS of the crop. Moreover the nitrogen stress detected during the initial growth stages 35-45 DAS of the crop are most useful for top dressing dose of nitrogen to overcome the stress. Using stepwise regression approach nitrogen deficiency model was developed, which showed NDRE is the only highly significant (P<0.01) index for detecting nitrogen deficiency in wheat crop.

Keywords

• Nitrogen stress
• NDRE
• NDVI
• spectral signature
• stepwise regression model, wheat