Submitted to: Meeting Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 9/9/2008
Publication Date: 11/1/2008
Citation: Iqbal, J., Read, J.J., Thomasson, A.J. 2008. Site specific N application and remote sensing of cotton crop. Proceedings International Conference on Advances in Space Technologies. Islamabad, Pakistan, November 29-30, 2008. 6 p.
Interpretive Summary: Variability in cotton lint yield across a field can be caused by natural spatial variability in soil properties, field elevation and/or topography, and sometimes man-made variability from an intended or unintended management practice. Recent advances in agricultural technologies that rely on the global positioning system (GPS), remote sensing, GPS-equipped yield monitor, and analysis of these data using a Geographic Information System (GIS), has created supporting industries in many regions of the world. Two valid reasons for using such technologies include the ability to 1) apply agrochemical more efficiently according to the potential of the soil or need of the crop, and 2) reduce environmental impact of non-point source pollution from agricultural activities. Research suggests site-specific application is profitable if the field has enough spatial variability in soils or topography. This study determined the response of cotton lint yield to three N rates in each of two field transects and variability in soil properties. Crop health was estimated from periodic measurements of crop reflectance (NDVI). Cotton NDVI values in July were used to create a map of predicted cotton yield, that differed by about 213 kg/ha from the measured yield for the field. Results indicated NDVI was sensitive to changes in N rate and specific soil properties, and the most productive sites had soil textural classes of silty clay loam or silt loam.
Technical Abstract: A spatial variable nitrogen (N) rate trial and remote sensing of cotton crop was conducted during 2003 at Paul Good Farms, Mississippi, USA. The N rate trial consisted of three, 8-row transects at the east and west side of the field that were selected to represent variable soil and elevation features across the field. The three N treatments (UAN, urea-ammonium nitrate solution) were 0, 112 and 185 kg N/ha. Multispectral aerial images of normalized difference vegetation index (NDVI) were collected with a 0.5 m spatial resolution of cotton crop throughout the growing season and resulting mean values were registered to 52 sites (n=16, a 2x2 m area of interest) in each transect. The statistical analysis showed that the highest cotton lint yield was obtained from rows and sites provided 185 kg N/ha, followed by 112 and 0 kg N/ha treatments. Cotton NDVI values in July were used to create a map of predicted cotton yield that differed by about 213 kg/ha from the measured yield for the field. Results indicated NDVI was sensitive to changes in N rate and specific soil properties, and the most productive sites had soil textural classes of silty clay loam or silt loam.