MAGNITUDE AND PERSISTENCE OF HERBICIDE RESIDUES IN FARM DUGOUTS AND PONDS IN THE CANADIAN PRAIRIES

Farm ponds or dugout waters were monitored for residues of seven major herbicides used in the Canadian prairies from fall of 1987 to spring of 1989. The frequencies of confirmed detection of herbicides in water samples, depending on the time of sampling, in decreasing order were: 2,4-dichlorophenoxyacetic acid (2,4-D; 93-100%), diclofop (46-95%), b romoxynil(50-85%), 4-chloro-2-methyl-phenoxyacetic acid (MCPA; 33-70%) , triallate (28-63%), dicamba (17-55%), and trifluralin (0-18%). The c orresponding frequencies of quantifiable residues (greater than or equ al to 0.05 mu g/L) were lower, ranging from 75 to 86% for 2,4-D to 0 t o 7% for dicamba. Median residues in all water samples were near or be low the quantification limits of 0.05 mu g/L. Maximum residues varied widely and were (mu g/L): trifluralin (not detectable [ND]-0.11), brom oxynil (0.27-0.33), dicamba (ND-11.2), triallate (0.05-0.87), MCPA (0. 12-1.97), 2,4-D (0.64-2.67), and diclofop (0.27-3.47). Maximum residue s were seasonal and declined to near or below detection limits by the following sampling time. Median values were two to three orders of mag nitude less than the corresponding maximum allowable concentration and interim maximum allowable concentration guidelines for drinking water in Canada and the United States. Maximum values were also less than t hese guidelines. Only the maximum values for residues of MCPA and 2,4- D approached the guidelines for these herbicides in water used for irr igation.