Tuesday, November 14, 2006

Optimal Nitrogen Fertilizer Rate for Corn as a Function of Terrain Attributes and Ground Based Sensor Variability.

Megan Poulette1, Thomas Mueller2, Scott A. Shearer1, and Paul Cornelius1. (1) Univ of Kentucky, 101 T. H. Morgan Building, Lexington, KY 40506-0225, (2) University of Kentucky, U. of Kentucky/Plant & Soil Sci., N122G Ag. Science Center North 0091, Lexington, KY 40546-0091

It may be possible to predict the occurrence of or detect N stress using precision agriculture technologies prior to side dressing corn with fertilizer N.  The objective of this study was to determine whether N response was related to terrain data, electrical conductivity, and soil type variability, and to determine whether N stress could be detected with spectral measurements.  Four N rates were applied in randomized complete block design in central Kentucky fields over a three year period.  Corn grain yield was measured with a mass flow sensor, soil EC data was measured by direct contact, crop reflectance was measured with multispectral and hyperspectral scanners, and elevation data was collected with survey grade GPS.  Digital elevation models were generated and used to calculate terrain attributes.  Analysis of covariance was used to determine whether there were interactions between yield and site-specific indices.  In some fields, results indicated that the response to nitrogen fertilizer for corn changed for different levels of slope, NDVI, and EC.  The optimum N rate varied in a way that indicated that EC and terrain attributes may be useful for the development of variable rate prescription maps, and that hyperspectral data could potentially be a useful tool in the prediction of optimal N rate.

Handout (.pdf format, 193.0 kb)