No abstract

Herbicide drift is the movement of herbicide away from its intended target. The effect of drift on non-target plants is considered in environmental risk assessments, where the goal of the assessment is to protect plant populations and communities. The aim of this study was to evaluate the assumption that the single species tests used in risk assessments are fully protecting wild plant populations, as they do not account for interspecific interactions. In a greenhouse two-species competition experiment, it was found that the competitive interactions between the model species, Centaurea cyanus and Silene noctiflora, were affected by low doses of glyphosate representing drift. These changes could affect both of their populations in the long-term, and would not be detected using current test guidelines. As interspecific competition is an important determinant of plant community structure, competitive interactions may need to be included in risk assessment to make more credible predictions on the effects of herbicide drift on non-target plants.

Plant competitive interactions influence the effect of herbicides, and the effect of competitive interactions on plant responses may be important to include in the ecological risk assessment of herbicides. In the present study the effect of competitive interactions and sublethal doses of 2 herbicides on plant species was investigated in competition experiments and fitted to empirical competition models. Two nontarget species commonly found in agroecosystems (Centaurea cyanus L. and Silene noctiflora L.) and 2 herbicides (glyphosate and metsulfuron methyl) were used in separate experiments. Plants were sprayed at the 6‐ to 8‐leaf stage. Effects of herbicide treatments and plant density were modeled by generalization of a discrete hyperbolic competition model. The 10% effective dose (ED10) was calculated for C. cyanus. All experiments showed that as density increased, plants were negatively affected. Furthermore, in all cases, C. cyanus remained a better competitor than S. noctiflora. Nevertheless, the density of S. noctiflora (competitor) was an influential element in determining the ED10 of C. cyanus measured at the mature stage. With herbicide exposure, the competitive interactions were further altered; C. cyanus was less affected by glyphosate when S. noctiflora increased to high density. In contrast, at the young stage, conspecific density was important in determining the sensitivity of C. cyanus to metsulfuron methyl, whereas the density of the competitor S. noctiflora had a limited influence. Overall, the results demonstrate the importance of integrating the effect of herbicide and species interactions measured at the reproductive stage into the ecological risk assessments of pesticides. Environ Toxicol Chem 2019;38:2053–2064. © 2019 SETAC.