Aquatic herbicide diquat threatens native plants at ecologically relevant doses

In this post, Verena Sesin talks about their recent publication “Macrophytes are highly sensitive to the herbicide diquat dibromide in test systems of varying complexity”.

Diquat dibromide is applied in North American waterbodies to control nuisance macrophytes. These are mostly non-native species that have become invasive by spreading widely and outcompeting native species. Herbicide application has been approved in affected waterbodies, as it is thought to pose minimal long-term risks to non-target aquatic biota. However, its effect on non-target native macrophytes is less clear. Diquat dibromide could affect a range of native species growing in the waterbody, because it is a non-selective, highly water-soluble contact herbicide.

The test macrophytes Elodea canadensis, Myriophyllum spicatum, Ceratophyllum demersum and Hydrocharis morsus-ranae (from left to right) (photo by V. Sesin)

Here we determined the sensitivity of both native and non-native macrophytes grown singly and in mesocosms to diquat dibromide. We exposed Elodea canadensis Michx. (Canadian waterweed), Myriophyllum spicatum L. (Eurasian watermilfoil), Ceratophyllum demersum L. (Coontail) and Hydrocharis morsus-ranae L. (European frogbit) to a range of diquat dibromide concentrations (4.7 – 1153 µg/L), corresponding to 0.4 – 100% of the recommended label rate of the formulated product. The mesocosm experiment contained all four plant taxa in the same system along with caged amphipods, tadpoles, phytoplankton and periphyton.

Mesocosm set-up (photo by V. Sesin)

In both single species and mesocosm tests, we detected severe direct effects of diquat dibromide on macrophytes, with almost 100% mortality at 74 µg/L. The most sensitive species in the single species tests, E. canadensis, showed almost 100% mortality at concentrations below the detection limit of 5 µg/L. Effects occurred very rapidly and showed no difference in severity between native and non-native macrophytes or complexity of test systems.

These results suggest that diquat dibromide could be applied at a considerably lower label rate, while still achieving effective control of nuisance macrophytes. In the scope of a future re-evaluation, regulatory authorities could adjust the current label rate taking into account variations in depth, volume and hydrology of the aquatic application sites. Moreover, authorities could propose maximum acceptable diquat dibromide concentrations in water following macrophyte control. For the underlying risk assessment process, authorities could request and evaluate data from a suite of macrophyte species, combining toxicity data from lower-tier single species tests and higher-tier tests such as mesocosm studies. This would account for indirect and long-term effects and increase environmental realism and predictive ability.

This research was conducted as MSc thesis for the international “Master of Science program in Ecotoxicology” of University of Koblenz-Landau, supervised by researchers of the National Wildlife Research Centre of Environment and Climate Change Canada in Ottawa. The work was performed in collaboration with the University of Ottawa and Carleton University, and funded by Fisheries and Oceans Canada.

The paper authored by Verena Sesin, Rebecca L. Dalton, Céline Boutin, Stacey A. Robinson, Adrienne J. Bartlett and Frances R. Pick was published in Ecotoxicology and Environmental Safety. The article can be accessed online for free until October 30, 2018.

You may also be interested in:

42 days set-up, 42 days mesocosm study