SEASONAL-VARIATION IN HERBICIDE LEVELS DETECTED IN SHALLOW ALBERTA GROUNDWATER

A three-year study was conducted at Lethbridge, on a 1-ha field (clay loam soil) with a history of previous herbicide applications, to deter mine herbicide levels in shallow (0.3-5.8 m) groundwater and the effec t of seasonal events such as herbicide application and rainfall/irriga tion. A mass-selective detector, gas chromatographic method was used t o analyze for seven different herbicides. Samples were collected from 6-m monitoring wells at various intervals (9-32 d) after herbicide app lications made in 1991-1993. 2, 4-D (0.05-5.2 ppb), bromoxynil (0.03-8 .4 ppb), diclofop (0.1-11 ppb), MCPA (0.2-0.3 ppb) and triallate (0.1- 0.5 ppb) were detected in 17-61% of the wells sampled. Herbicide level s depended upon the interval and the timing and amounts of rainfall/ir rigation after spraying. All levels were below the Canadian drinking w ater guidelines except for one detection of bromoxyril (8.4 ppb) and d iclofop (11 ppb). We concluded there are large seasonal and spatial va riations in herbicide levels in shallow Alberta groundwater. Levels ar e highest when the first moisture event after a herbicide application is a heavy rainfall/irrigation ('worst case' scenario). Levels are low er when the applied herbicides are first 'set-in' by gentle rains, the n a drying period occurs and the residues adsorb to the soil. Subseque nt heavy rainfall/irrigation does not flush large amounts of these ads orbed residues into the groundwater. This latter 'best case' scenario probably occurs more often than the former 'worst case' scenario.