Picloram Movement in Soil Solution and Streamflow from a Coastal Plain Forest

Picloram (4‐amino‐3,5,6‐trichloropicolinic acid) was aerially applied to a longleaf pine ( Pinus palustris L.) site in the upper coastal plain of Alabama to control kudzu [ Pueraria lobata (Willd.) Ohwi]. Pellets (10% a.i.) were spread at the rate of 56 kg ha −1 on loamy sand Typic Kanhapludult soils. Movement of this herbicide was monitored with mineral soil samples, tension‐cup lysimeters, flow‐proportional streamflow samplers, and discrete samplers. Picloram levels in the upper 15 cm of mineral soil peaked at 0.96 to 2.25 mg kg 1 25 d after application, depending on slope position, and declined to 0.13 to 0.29 mg kg 1 1 yr later. In soil solution, picloram was detected at a depth of 0.4 m between 26 and 273 d after application. Only 4 of 15 lysimeters consistently contained detectable residues. Maximum picloram levels in soil solution were 130, 450, and 191 mg m −3 for ridge, midslope, and toe‐slope positions, respectively. Downstream monitoring began 4 d after the herbicide application, and an initial concentration of 68 mg m −3 of picloram was detected. The maximum downstream concentration of 77 mg m 3 occurred 18 d after the application, immediately after the second storm event. Downstream levels dropped to <10 mg m 3 after 90 d and to <2 mg m −3 after 200 d. Following localized retreatment along the stream more than a year after the initial treatment, levels climbed again into the 20 to 30 mg m 3 range. Most of the initial off‐site movement came from a perennial stream that had been inadvertently treated, but subsequently storm runoff was the largest contributor to stream contamination. Picloram residues in this stream were similar to those observed downstream, but they were higher (up to 241 mg m 3 and dropped faster to below 2 mg m 3 after D 150.