Combination effects of herbicides on plants and algae: do species and test systems matter?

Risk assessment of herbicides towards non-target plants in Europe is currently based solely on tests on algae and floating aquatic plants of Lemna sp. Effects on terrestrial non-target species is not systematically addressed. The purpose of the present study was to compare combination effects of herbicide mixtures across aquatic and terrestrial test systems, and to test whether results obtained in the traditional aquatic test systems can be extrapolated to the terrestrial environment. This was done by evaluating ten binary mixtures of nine herbicides representing the seven most commonly used molecular target sites for controlling broadleaved weeds. Data were evaluated statistically in relation to the concentration addition model, and for selected concentrations to the independent action model. The mixtures were tested on the terrestrial species Tripleurospermum inodorum (L.) Schultz-Bip. (Scentless Mayweed) and Stellaria media (L.) Vill. (Common Chickweed), and on the aquatic species Lemna minor L. (Lesser duckweed) and the alga Pseudokirchneriella subcapitata (Korschikov) Hindak. For the two mixtures of herbicides with the same molecular site of action, the joint effect was additive. For the eight mixtures of herbicides with different sites of action, two of the mixtures were consistently antagonistic across species, while for the remaining six mixtures the joint effect depended on the species tested. This dependence was, however, not systematic, in the sense that none of the species or test systems (terrestrial versus aquatic) had a significantly higher probability of showing synergistic or antagonistic joint effects than others. Synergistic interactions were not observed, but approximately 70% of the mixtures of herbicides with different sites of action showed significant antagonism. Hence, the concentration addition model can be used to estimate worst-case effects of mixtures of herbicides on both terrestrial and aquatic species. Comparing the sensitivity of the species to a 10% spray drift event showed that the terrestrial species were more vulnerable to all herbicides compared with the aquatic species, emphasising the importance of including terrestrial non-target plants in herbicide risk assessment.