A MICROCOSM APPROACH TO THE DETECTION OF THE EFFECTS OF HERBICIDE SPRAY DRIFT IN PLANT-COMMUNITIES

It is often difficult to measure and predict the impacts of toxic chem icals, such as herbicides, on natural communities. This is especially true under conditions of spray drift when the amount received by the o rganisms downwind from the sprayer may be at sub-lethal doses. Laborat ory experiments are either artificial, or have not been generally carr ied out over long enough time periods, to be realistic. Field experime nts are often difficult because of the high variability inherent in na tural populations. Here an intermediate microcosm approach was used, w here standardized artificial communities (eight dicotyledons with and without a grass) were tested. The artifical communities included speci es typical of British woodland margins, hedgerows and field margins; c ommunities with a high conservation interest, yet potentially under th reat from spray drift. The microcosms were placed downwind of a spraye r and exposed to one of the following herbicides: glyphosate, mecoprop and MCPA. This approach ensures that the communities were standardize d at the start and have been exposed to realistic doses of herbicide. The experiments reported here were carried out for at least three year s with exposures to herbicides repeated each year. The effects of diff erential herbicide exposure downwind of the sprayer were measured on s pecies yield, flowering performance, seed production, seed viability a nd invasion by new species. Responses were extremely variable, but all species showed some effects in some years. Some patterns emerged. For example, one group of species appeared to be more successful near to the sprayer. This was particularly true of the grass when exposed to M CPA and mecoprop. The performance of most species was reduced under th e sprayer, and there was a general recovery with increasing distance d ownwind. A few species showed increased performance in the intermediat e downwind zone (2-4 m) and this may be due to a hormonal effect on gr owth processes, or an effect of reduced interference from other commun ity members. Generally there were few effects on seed production or se ed viability. An important result was that most effects were confined within an 8 m zone, as there were few significant differences between plants exposed at 8 m and those untreated. Although damaging effects w ere found in the immediate downwind zone from the sprayer, the restric tion of effects to 8 m suggests that a buffer zone of this size would be adequate to protect sensitive habitats from most deleterious impact s on community processes. (C) 1997 Academic Press Limited.