Systems Agronomic and Economic Evaluation of Transgenic and
Conventional Varieties in the Texas High Plains
February, 2001
Dr. Randy Boman, Extension Agronomist-Cotton
Mr. Mark Kelley, Extension Assistant-Cotton
Mr. Alan Helm, Extension Assistant-Cotton
Texas Agricultural Extension Service
Lubbock, TX
Materials and Methods
For scientific validity, three replications of each variety were included at each location. Plot size was of sufficient size to enable the compositing of all replications of each individual variety into a single module at harvest. Each individual variety had at least three acres total (approximately one acre per plot with three replications equals three acres total). A forced randomization was used at each location. This was a requirement due to the potential for drift of Buctril and Roundup Ultra to adjacent non-herbicide tolerant varieties. For example, the Roundup Ready varieties were planted in a contiguous block, with a fill variety between the next herbicide system. Varieties within the next herbicide system were then planted, with a fill variety between the last system. Varieties were randomized in each replication and herbicide system, but the forced randomization due to herbicide system was maintained. All fill varieties were treated with conventional herbicides and were not used for data acquisition.
Preplant incorporated and/or preemergence herbicide applications were made at the discretion of the producer-cooperator. Broadcast over-the-top herbicide applications were made using project equipment and project personnel or by the cooperator. Post-directed herbicide applications were made by the producer-cooperator with the guidance of project personnel. Weed species spectrum and densities was determined by project personnel working with the cooperator. Control of weed escapes (hoeing and/or spot spraying) was performed by project personnel and records were kept by herbicide system to facilitate economic analysis.
Plant mapping data were derived from plant mapping 6 representative plants per plot on each mapping date. Plot weights were determined at harvest using a boll buggy with integral electronic scales and grab samples were obtained from each plot. Modules were followed through the ginning process to determine lint turnout, USDA fiber quality, and loan value. Ginners were asked to gin each module separately and to tie off any remnant bales obtained in the ginning process in order to determine more precisely the turnout and lint yields. Data were then converted to a per acre basis and appropriate statistical analyses performed.
Three producer-cooperator locations were utilized for this project. Those locations included medium, high, and low weed control input sites.
