Relationships between cotton nitrogen status and sensor based reading

Authors: Balkcom, Kipling; SHAW, JOEY - Auburn University; RODEKOHR, DONN - Auburn University; FULTON, JOHN - Auburn University

Submitted to: National Cotton Council Beltwide Cotton Conference
Publication Type: Proceedings
Publication Acceptance Date: 1/12/2011
Publication Date: 4/25/2011
Citation: Balkcom, K.S., Shaw, J., Rodekohr, D., Fulton, J.P. 2011. Relationships between cotton nitrogen status and sensor based reading. In: Boyd, S., et al., editors. Proceedings of the 2011 National Cotton Council Beltwide Cotton Conference, January 4-7, 2011, Atlanta, Georgia. p.1356-1359.

Interpretive Summary: Previous research has shown that cotton (Gossypium hirsutum L.) yields can vary across the landscape based on the management system utilized; however, typical N fertilizer rates for cotton are applied uniformly across the field. Sensor technology previously utilized in other crops can potentially quantify cotton N status and relate sensor output to an N recommendation that will vary across the field and simultaneously improve cotton yields and profitability. An experiment was initiated by scientists from the National Soil Dynamics Laboratory and Auburn Univ. in the fall of 2007 at the E.V. Smith Research Center, Field Crops Unit near Shorter, AL that utilized two soil management systems (conventional-CT and conservation-NT) and 4 N rates (0, 40, 80, 120 lb/ac) across a 20 ac. field. At pre-determined sampling points (12 x 60 ft. grids stratified within each management zone) 25 upper-most mature leaves were collected along with plant heights, sensor measurements, and 3.28 ft. of whole plant biomass from the same pre-determined locations across landscape positions around 1st square and mid-bloom. Seed cotton yield was determined across the field with a spindle picker equipped with GPS and yield monitor. Although NDVI was well correlated with various plant parameters, no clear relationship was identified between NDVI sensor readings and lint yield for the 2008-2010 growing season.

Technical Abstract: Previous research has shown that cotton (Gossypium hirsutum L.) yields can vary across the landscape based on the management system utilized; however, typical N fertilizer rates for cotton are applied uniformly across the field. Sensor technology previously utilized in other crops can potentially quantify cotton N status and relate sensor output to an N recommendation that will vary across the field and simultaneously improve cotton yields and profitability. An experimental site was established in a 22 ac. field in the Coastal Plain consisting of Typic and Aquic Paleudults that had management systems in place for 9 years. A split plot design was utilized with main plots consisting of two soil management systems (conventional-CT and conservation-NT) and split plots of 4 N rates (0, 40, 80, 120 lb/ac) with six replications of 12-ft. strip-transects (4-row widths) across the field that intersect management zones. At pre-determined sampling points (12 x 60 ft. grids stratified within each management zone) 25 upper-most mature leaves were collected along with plant heights, sensor measurements, and 3.28 ft. of whole plant biomass from the same pre-determined locations across landscape positions around 1st square and mid-bloom. Seed cotton yield was determined across the field with a spindle picker equipped with GPS and yield monitor. Although NDVI was well correlated with various plant parameters, no clear relationship exists between NDVI/Cum. GDD and lint yield for the 2008-2010 growing season.