Donnie K. Miller, Robert G. Downer, Steve T. Kelly and P. Roy Vidrine

Transgenic technology, where genes conferring resistance to certain herbicides or protection against select insect species are transferred to plants, has had a dramatic effect on cotton production. According to National Cotton Council figures for 2002, more than 70 percent of Louisiana cotton acreage was planted to transgenic varieties. Cotton varieties resistant to herbicides glyphosate (Roundup Ready), bromoxynil (BXN) and glufosinate (Liberty Link) have been developed. Roundup Ready and BXN cotton have been commercially available for several years; Liberty Link varieties are to be available on a limited commercial basis in 2003. Use of this new technology creates the possibility for off-target movement, known as drift, onto nonherbicide-resistant cotton planted in close proximity or for misapplication from sprayer contamination.

To assess possible negative effects, research was conducted at the Northeast Research Station in St. Joseph, La., in 2000 and at the Macon Ridge Research Station in Winnsboro, La., in 1999 and 2000. The methodology involved testing reduced rates of glyphosate (Roundup Ultra formulation), bromoxynil (Buctril) and glufosinate (Liberty) on two different cotton varieties (Stoneville 474 at Macon Ridge and DP33B at Northeast) at three different growth stages (two-node, five-node and nine-node). Cotton response was assessed in seedcotton yield as well as in the following: injury 14 days after treatment, plant height 30 days after treatment, dry weight 30 days after treatment, nodes above white flower, percentage open bolls and final plant population.

Roundup Ultra

Based on visual assessment, cotton was more tolerant to glyphosate at the nine-node growth stage. Plant dry weight was reduced with glyphosate rates of 2 ounces per acre or higher applied at the two- and five-node growth stages in two of three experiments. Weights were not affected by glyphosate at the nine-node stage. Plant height was also unaffected by lower rates, but reduction was noted for all growth stage by experiment combinations with the exception of nine-node for both experiments in 2000 with increase in herbicide rate of 2 ounces per acre or higher. Cotton maturity delay, as noted by increase in node above white flower (NAWF) number, was observed only at the highest glyphosate rate applied to two- and five-node cotton in one of three experiments. Percent open boll data analysis indicated a decrease in chance of observing an open boll with increasing glyphosate rate, and this effect was higher at the five-node stage compared to the two- and nine-node stages in two of three experiments. Although seedcotton yield was reduced 9 percent to 13 percent for the highest rate when compared to lower rates, yield following glyphosate application was equivalent to that for the nontreated control.

Buctril

Based on visual assessment, cotton response was reduced as application timing was delayed from two-to five-node in all experiments and from five- to nine-node in two of three experiments. Plant height reduction response decreased with increasing growth stage at time of application. Plant dry weight was most negatively affected following application at the two-node stage. Bromoxynil application, based on node above white flower number (NAWF), did not result in maturity delays but did promote earliness when applied at 4 ounces per acre to two- and five-node cotton in one of three experiments. Final plant population was reduced only at the two-and five-node timings, with response more pronounced at the initial timing. Seedcotton yield following bromoxynil application at the highest rate to two-leaf cotton was reduced 34 percent compared with other rates and the nontreated control, which were equal. Bromoxynil applied to five- or nine-node cotton did not reduce yield significantly.

Liberty

Based on observation, cotton response decreased as application timing was delayed in one of three experiments. Injury response was slightly increased with application at the five compared to two-node growth stage and was not significant for the latest application timing (nine-node) in two of three experiments. In two of three experiments, plant height reduction response was lowest at the five-node stage and highest at the nine-node timing. Regardless of application timing, plant dry weight was affected negatively only with the highest rate of glufosinate. Glufosinate application, based on node above white flower number and percent open boll, did not result in a maturity delay. Final plant population was reduced in all experiments at the two-node application and in one of three experiments at the five-node timing. Glufosinate application did not affect final plant population adversely when applied to nine-node cotton. Negative effects on cotton growth were not manifested in seedcotton yield reduction following glufosinate application.

Prevention Is Best

Although various negative effects on growth were observed following application of Roundup Ultra, Buctril or Liberty herbicides at rates simulating those experienced with drift and sprayer contamination, cotton was able to recover and yields were equal to nontreated plants in almost every case. Yield reduction was observed only with Buctril at the highest rate (4 ounces per acre) applied to cotton in the least advanced growth stage (2-node). Although the cotton was able to recover fully, it should be noted that late-season growing conditions were optimum at all locations because of supplemental irrigation at Winnsboro and adequate rainfall at St. Joseph. Cotton was managed for insect pests throughout the season. Had environmental conditions been less than optimal, the growing season shortened or insect damage more extensive, cotton may not have been able to recover fully, and detrimental effects from herbicide application may have caused yield reduction. In addition, rates observed in these studies are representative of those that would be expected in drift or sprayer contamination situations, and higher rates may result in more deleterious effects. Therefore, producers are cautioned to take all measures to avoid herbicide drift and sprayer contamination.

Donnie K. Miller, Associate Professor, Northeast Research Station, St. Joseph, La.; Robert G. Downer, Assistant Professor, Department of Experimental Statistics, Baton Rouge, La.; Steve T. Kelly, Assistant Specialist (Weed Science), Scott Research, Extension, and Education Center, Winnsboro, La.; and P. Roy Vidrine, Professor, Dean Lee Research Station, Alexandria, La.

(This article was published in the 2003 spring issues of  Louisiana Agriculture magazine.)