LINKING HERBICIDE DISSIPATION TO SOIL ECOLOGICAL RISK ALONG RIGHT-OF-WAYS IN THE YUKON TERRITORY, CANADA

Abstract

In the Yukon Territory, vegetation management along transmission right-of-ways (ROWs) is conducted using brushing and mowing techniques alone. When cut, target species, such as Populus spp. and Salix spp., grow rapidly, shortening maintenance cycles. Long-term vegetation management may be improved by integrating herbicide application. However, prior to implementation, the dissipation and toxicity of herbicides in northern latitudes needed to be assessed. The dissipation of Garlon XRT (triclopyr) and Arsenal Powerline (imazapyr) in soils was assessed at five ROW locations representative of the main ecoregion types where ROWs occur within the Yukon Territory. Soils from four sites were collected to a depth of three centimetres at 1, 30 and 365 days after treatment (DAT) to determine dissipation of herbicides for each of three application methods (cut stump, point injection and backpack spraying). Soils from a fifth site were collected more frequently on day 0, 1, 3, 7, 14, 21, 30 and 60 DAT to better determine the dissipation time of each herbicide in Yukon Territory soils. Mean triclopyr concentrations at 365 DAT were 0.01 ± 0.01 mg ai kg-1 and 0.24 ± 0.20 mg ai kg-1 for cut stump and point injection, respectively. Whereas, the mean concentrations for imazapyr cut stump and point injection treatments at 365 DAT were 0.01 ± 0.002 mg ai kg-1 and 0.03 ± 0.02 mg ai kg-1, respectively. Dissipation rates for the backpack spray treatment indicated that triclopyr (time to 50% of the initial concentration [DT50] of 1 DAT) dissipated faster than imazapyr (DT50 of 16 DAT). Residues from the cut stump and point injection treatments dissipated considerably between 30 and 365 DAT for both herbicides. Soil dissipation data was linked to a series of standardized soil toxicity tests, including three soil invertebrate tests (Enchytreaus crypticus, Folsomia candida, and Oppia nitens) and three soil enzyme tests (arylsulfatase, B-glucosidase and phosphatase). Expected maximum application concentrations (75.5 mg triclopyr kg d.w.-1 and 12 mg imazapyr kg d.w.-1) were below the 28-day (28-d) EC25 for all species tested. Even sensitive endpoints such as 28-d LC10 and 28-d EC10 were generally above the expected application concentrations. E. crypticus and F. candida reproduction endpoints were often more sensitive to triclopyr when compared to imazapyr in the soils tested. Soil enzymatic activity could not be adequately modelled for dose response. However, for both the invertebrates and soil enzymes tested, clear site differences occurred in response to habitat quality specifically related to soil pH and total organic carbon. Weight of Evidence (WOE) and Toxic Exposure Ratios (TER) were used to characterize the risks associated with herbicide application in northern latitudes providing both qualitative and quantitative means to effectively communicate the results to the public. In this study the WOE approach demonstrated that potential environmental concentrations were below not only the effective concentration at 25% (28-d EC25), but also the effective concentration at 10% (28-d EC10) values for all invertebrate species tested. While the TER approach identified that some ecological risk was present to soil organisms with the use of triclopyr, no unacceptable risks were identified through the application of imazapyr. The identified risks of triclopyr application are close to the critical trigger value of five and it is likely that soil invertebrate communities would recover less than one year after application.