Determination of buffer zones to protect seedlings of non-target plants from the effects of glyphosate spray drift

Abstract There is an increasing need to protect semi-natural vegetation from the potential effects of herbicide drift. One way to protect sensitive sites is to surround them with a no-spray buffer zone. Earlier estimates of buffer zone size based on bioassay experiments with established perennials suggested zones needed to be 6–10 m wide. In this paper four bioassay experiments are reported, where seedlings grown in trays were exposed downwind of glyphosate applications and taken to a glasshouse for assessment. Three experiments were done with Lychnis flos-cuculi seedlings including one with different surrounding grass structures, and Experiment 4 tested the response of 15 species typical of semi-natural vegetation. The mortality of Lychnis flos-cuculi varied between experiments and appeared more or less unaffected by grassland structure except immediately downwind of the sprayer. The multi-species experiment indicated a wide sensitivity to spray drift, and one species was affected between 15 and 20 m downwind. Thus, seedlings of some species were affected at greater distances than established plants, indicating either greater capture of drift or a greater sensitivity. On sites where seedling establishment is an important mechanism for community regeneration, buffer zones may need to be 20 m wide.