L.L. Lyon, J.W. Keeling, T.A. Baughman, T.S. Osborne, and P.A. Dotray; Texas Agricultural Experiment Station, Lubbock, Texas Cooperative Extension, Vernon, Oklahoma State University, Altus, Texas Tech University, Lubbock, and Texas Cooperative Extension, Lubbock.

ABSTRACT

In 2001, approximately 70% of the 15.5 million acres of upland cotton planted in the United States was glyphosate tolerant.  The potential for herbicide drift or misapplication exists on non-glyphosate tolerant cotton cultivars that are often planted adjacent to glyphosate tolerant cotton.  These include conventional, non-transgenic varieties, as well as other non-glyphosate tolerant transgenic varieties. Experiments were conducted at three locations in 2002, including Lubbock, TX; Munday, TX; and Altus, OK to determine the effects of low rates of glyphosate (similar to drift) on non-glyphosate tolerant cotton.    At the Lubbock location, Paymaster HS26 was planted and at Munday and Altus, DPL 237B was planted.  Glyphosate (Roundup UltraMax) was applied at 0.38 lb ae/A, 0.19 lb ae/A, 0.094 lb ae/A, 0.047 lb ae/A, and 0.023 lb ae/A (1/2X, 1/4X, 1/8X, 1/16X, and 1/32X of 0.75 lb ae/A, respectively) postemergence-topical to cotton at the cotyledon to 1-leaf (COT-1 lf), 4- to 5-leaf (4-5 lf), pinhead square (PHSQ), and first bloom (FBLM) growth stages.  An additional treatment of glyphosate at 0.75 lb ae/A was applied at all growth stages at the Lubbock and Munday locations.  Cotton visual injury ratings were taken at 14 days after treatment (DAT), 21 DAT, 28 DAT, and at the end of the season.  Plants were mapped at the end of the season and cotton lint yields and quality were determined.

At Lubbock, 14 DAT, injury from 8 to 98% was observed from rates > 0.023 lb ae/A when applied at COT-1 lf, 4-5 lf, and PHSQ.  The lowest rates (0.023 and 0.047 lb ae/A) did not show any visual injury when applied at FBLM compared to higher rates, which injured cotton 60 to 95%.  Injury decreased slightly from higher rates applied early season (COT-1 lf and 4-5 lf), with only rates > 0.19 lb ae/A showing 10 to 82% injury by the end of the season.  Later season (PHSQ and FBLM) applications however, showed an increase in injury from the 0.38 and 0.75 lb ae/A rates to at or near 100% visual injury.  Glyphosate applications > 0.094 lb ae/A applied at PHSQ caused visual injury, while rates > 0.047 lb ae/A applied at FBLM injured cotton.  Yield was only reduced by rates > 0.19 lb ae/A applied at COT-1 lf, although all but the 0.023 lb ae/A application showed visual injury 14 DAT.  Yield was reduced by all glyphosate rates except 0.023 lb ae/A on 4-5 lf cotton.  All rates applied at PHSQ reduced cotton yield, even 0.023 lb ae/A, which showed no visual injury 14 DAT.  Only glyphosate at 0.023 lb ae/A applied at FBLM did not reduce yields.  No yield was produced when 0.75 lb ae/A of glyphosate was applied at PHSQ and FBLM or from the 0.38 lb ae/A rate applied at FBLM.

At Munday, 14 DAT, 20 to 82% injury was observed from all glyphosate rates applied at COT-1 lf, and rates > 0.023 lb ae/A applied at 4-5 lf cotton caused visual injury (30 to 81%).  Glyphosate rates > 0.094 lb ae/A applied at PHSQ caused 40 to 70% visual injury, while 12 to 30% injury was observed from rates > 0.19 lb ae/A applied at FBLM.  The injury from FBLM applications was much less than injury at other growth stages.  By the end of the season, injury was only seen from 0.75       lb ae/A applied at the COT-1 lf and 4-5 lf stages.  Injury from glyphosate rates > 0.094 lb ae/A applied at the PHSQ stage was still apparent by the end of the season, but was reduced.  Glyphosate rates > 0.094 lb ae/A applied at FBLM showed visual injury at the end of the season.  Although visual injury decreased by the end of the season, yield reductions were observed with rates > 0.094 lb ae/A applied at all but the FBLM growth stage.  Even though all the glyphosate rates showed initial and end of season visual injury less than the injury observed at other stages, only 0.023 lb ae/A applied at FBLM did not decrease yield.

Glyphosate rates > 0.023 lb ae/A applied at the COT-1 lf and 4-5 lf stages showed visual injury 14 DAT at Altus (10 to 92%).  All rates applied at PHSQ injured cotton 10 to 40%.  Only glyphosate at 0.19 lb ae/A applied at FBLM showed any visible injury at 14 DAT (10%).  By the end of the season, the only visual injury observed was from 0.38 lb ae/A made at the COT- 1 lf (20%) and 4-5 lf (28%) stages and from 0.19 lb ae/A applied at FBLM (10%).  Yield was reduced at all growth stages from glyphosate applied at the 0.38 lb ae/A rate.  Glyphosate applied at 0.19 lb ae/A at the COT-1 lf and FBLM growth stages reduced cotton yields, while rates > 0.094 lb ae/A applied at 4-5 lf decreased yields.

Application timing and glyphosate rate affected cotton injury levels at all locations; however, visual injury did not always result in a yield reduction.  Yield loss tended to be over estimated by visual injury, especially for early applications (COT-1 lf, 4-5 lf).  Later applications (PHSQ, FBLM) affected yield more than early applications.  Cotton injury varied by location and was dependent on the growing season and crop conditions.