Water quality following application of picloram, triclopyr, and tebuthiuron on south Texas rangeland

Abstract

Current public perceptions of herbicide effects on water quality make it important to document them wherever possible. Objectives of this study were to determine if herbicides used for brush control influence: 1) vegetation composition, 2) amount of runoff, sediment, and nutrients produced, and 3) amount of herbicide in runoff. Research was conducted on the Senesa Ranch located 56 km west southwest of San Antonio, Texas, in the northern Rio Grande Plains resource region. Nine watersheds were instrumented to collect samples for water quality analysis, stage recorders monitored depth and duration of runoff, and vegetation cover was monitored using three 3O m line transects within each watershed. Three watersheds were treated with an aqueous solution of picloram and triclopyr (0. 56 kg ha-1) in October, 1992, three with a pelleted formulation of tebuthiuron (2.2 kg ha-1) in March, 1992, and three left untreated. Both picloram/triclopyr and tebuthiuron applications were found to alter vegetation composition. For the 30 month period of the study brush cover declined 28% and 40%, with an increase in perennial grass cover of 230% and 285% on the piclorarn/tticlopyr and tebuthiuron treatments, respectively. Herbicide applications caused a significant decrease in runoff and production of water quality parameters. On the picloram/triclopyr watersheds, runoff decreased 79%, sediment production 89%, and nitrates 70%. On the tebuthiuron watersheds, runoff decreased 38%, sediment production 76%, and nitrates 97%. Water samples were analyzed for picloram and triclopyr. Herbicide concentrations during the first runoff event 23 days after application were 151 and 142 mg L-1 for picloram and triclopyr respectively. These levels are substantially below 500 mg L-1, the allowable picloram level set for drinking water standards by the United States EPA. Within eight months picloram levels declined to less than 5 mg L-1, and triclopyr was undetectable (<2 mg L-1). By 18 months picloram levels were at or below detection limit ( 1mg L-1). Results indicate that vegetation response to herbicides utilized for brush control reduce water yield in the form of runoff, and may reduce the potential for non-point source pollution from South Texas Rangeland.