Abstract
Vegetative filter strips reduce herbicide runoff from cultivated fields owing to the ability of vegetation to delay surface runoff, promote infiltration, and adsorb herbicides. Previous research has shown that the annual runoff of each herbicide is typically less than 1 g ha−1. A model for the detection of the return period of rainfall events was recently proposed for a site in the north-eastern Po Valley, Italy. The return period model suggested that most of the herbicide loss by runoff (about 98%) is caused by a few, or even just one, extreme rainfall event with a return period of about 25–27 years, whereas ordinary events (4–5 each year) account for the rest. The present study aims to validate that model by comparing model predictions with the experimental results obtained in the 2010–2011 sampling season (independent test data), and to evaluate the effectiveness of the VFS. In addition, a 7-yr dataset of metolachlor and terbuthylazine concentration in real runoff events is summarised in order to highlight the medium-term magnitude of the pollution. Results show that on the Po Valley plain, 3–4 runoff events of low intensity are expected in spring–summer and that the consequent annual runoff of the herbicides metolachlor and terbuthylazine is about 0.5–0.7 g ha−1 yr−1. A summary shows that, owing to their chemical-physical properties, concentrations of the two herbicides are similar, both varying from about 0.01–300 μg l−1, with a potential pulse-like exposure risk for aquatic communities in waterways. This study showed that vegetative filter strips can reduce herbicide transfer to surface water by 90–98%, and should be suggested for environmental schemes at field and catchment scale.
Acknowledgements
This research was performed with financial support from MIUR, project PRIN 2008N7XHT5 ‘Ecotoxicological risk of plant protection products: ecological effects and mitigation measures’. We gratefully acknowledge ARPA Veneto and Dr. Francesco Rech for providing meteorological data and helpful suggestions for elaboration.