Testing Variable Rate Application (VRA) Advantages on Cotton

This research project was conducted by Lele Borghetti, a student at the Università degli Studi di Padova (Italy). He conducted this research at the University of Georgia Tifton Campus in 2011.


The project’s goal was to apply and evaluate the use of Variable Rate Application (VRA) on cotton. Cotton requires application of different agrochemicals such as defoliants and fertilizers. While defoliants facilitate harvesting operations, fertilizers assure an optimal plant growth. In particular, pre-planting fertilization guarantees good seed germination whereas mid-season side-dressing fertilization influences fruiting and final yield and so it is a key factor for cotton lint and seed production. Prediction of crop potential growth helps farmers to plan appropriate mid-season fertilizer applications resulting in reduction of cultivation costs and mitigation of fertilizer pollution (i.e. nitrate).

It is widely known that cotton is one of the most important crops in Georgia. Despite it having a long cropping history in this state, precision agriculture, and in particular variable rate application (VRA) of nitrogen, offers new opportunities to optimize crop cultivation. VRA is a precision agriculture technique that allows modulating agronomic inputs according to variability observed in the field. Basically there are two methods to manage variability: a) pre-developed application or prescription maps; b) real-time sense and treat systems. In general the first approach is more reliable because it allows you to verify collected data before making management decisions. However the development of proximal-sensing technologies is increasing the interest toward the real-time method which is the method farmers prefer because it requires less time.

The main aim of this study was assessing agronomic, economic and environment profitability of nitrogen VRA on cotton based on using the Normalized Difference Vegetative Index (NDVI) to measure variability in the growing cotton crop. NDVI is a vegetation index based on near infrared and red wavelengths reflected by plant leaves. The index is sensitive to leaf chlorophyll and so it is an effective indicator of plant biomass. A secondary goal was to evaluate the GreenSeeker spectroradiometer as a tool for managing VRA of cotton agrochemicals in Georgia.

The experiment was carried out in 2011 during the growing season. The experimental site was a 8-ha field located in Tift County. Soil variability was assessed through pre-planting soil analysis integrated by apparent soil electrical conductivity (ECa) measurements. Soil ECa is the capacity of the soil to transmit an electrical current. Since its values correlate with soil properties that affect crop productivity, ECa mapping provides a rapid and economic way to assess soil variability.

Proximal sensing was carried out during the crop season using six GreenSeeker® sensors mounted on the spray boom of a crop sprayer. The sensors measured NDVI. Maps of NDVI were obtained by connecting the GreenSeeker® user interface with a GPS. VRA nitrogen fertilization was based on prescription maps obtained by applying the Sensor Based Nitrogen Rate Calculators (SBNRCs) method developed by Oklahoma State University and modified by Clemson University to the NDVI map.

The prescription maps were developed in the Farm Works® software and then loaded into an Ag Leader Insight® variable rate controller. The Insight® controlled the rate of liquid side-dress nitrogen applied in response to the prescription map. Finally, a yield map was obtained at the end of the growing seasons with an Ag Leader cotton yield monitor mounted on a cotton picker.

NDVI mapThe results showed that VRA provided crop yields comparable to conventional fertilization while reduce the use of side-dress fertilizer and increasing nitrogen use efficiency. Variable nitrogen rate application seems to be a valuable and profitable technique to sustain cotton productivity and reduce the impact of agriculture on the environment.

The prescription maps were developed in the Farm Works® software and then loaded into an Ag Leader Insight® variable rate controller. The Insight® controlled the rate of liquid side-dress nitrogen applied in response to the prescription map. Finally, a yield map was obtained at the end of the growing seasons with an Ag Leader cotton yield monitor mounted on a cotton picker.

The results showed that VRA provided crop yields comparable to conventional fertilization while reduce the use of side-dress fertilizer and increasing nitrogen use efficiency. Variable nitrogen rate application seems to be a valuable and profitable technique to sustain cotton productivity and reduce the impact of agriculture on the environment.Yield map at the end of the growing seasons with an Ag Leader cotton yield monitor mounted on a cotton picker

Download Lele Borghetti’s thesis.