On the potential of on-line free-surface constructed wetlands for attenuating pesticide losses from agricultural land to surface waters

Abstract

Environmental contextPesticide losses from land to surface waters have the potential to cause ecological damage. Furthermore, pesticides in surface waters present a major challenge for water companies accessing these waters for the domestic supply, in terms of complying with water quality regulations. Here, we evaluate the potential of field- and ditch-scale free-surface constructed wetland systems for reducing pesticide transfers from land to surface waters. AbstractPesticides make important contributions to agriculture but losses from land to water can present problems for environmental management, particularly in catchments where surface waters are abstracted for drinking water. ‘On-line’ constructed wetlands have been proposed as a potential means of reducing pesticide fluxes in drainage ditches and headwater streams. Here, we evaluate the potential of two free-surface constructed wetland systems to reduce pesticide concentrations in surface waters using a combination of field monitoring and dynamic fugacity modelling. We specifically focus on metaldehyde, a commonly used molluscicide that is moderately mobile and has been regularly detected at high concentrations in drinking water supply catchments in the UK over the past few years. We also present data for the herbicide metazachlor. Metaldehyde losses from the upstream catchment were significant, with peak concentrations occurring in the first storm events in early autumn, soon after application. Concentrations and loads appeared to be minimally affected by transit through the monitored wetlands over a range of flow conditions. This was probably due to short solute residence times (quantified via several tracing experiments employing rhodamine WT – a fluorescent dye) exacerbated by solute exclusion phenomena resulting from patchy vegetation. Model analyses of different scenarios suggested that, even for pesticides with short aquatic half-lives, wetland systems would need to exhibit much longer residence times (RTs) than those studied here in order to deliver any appreciable attenuation. If the ratio of wetland surface area to the area of the contributing catchment is assumed to be a surrogate for RT (i.e. not accounting for solute exclusion), then model predictions suggest that this needs to be greater than 1% to yield load reductions of 3 and 7% for metaldehyde and metazachlor respectively.