Combination of passive and grab sampling strategies improves the assessment of pesticide occurrence and contamination levels in a large-scale watershed.

ABSTRACT

Fifty-one monitoring stations from the Water Framework Directive network (2000/60/CE) were selected in the Adour-Garonne basin (117,650 km2, SW France). These stations were characterized by a diversity of land use, implying different water pesticide contamination profiles. In each, Polar Organic Chemical Integrative Sampler (POCIS) deployment (14 days) and grab water samples (1 per period) were performed 6 times in 2016 in order to obtain contamination levels (29 pesticides monitored). The large amount of data collected during this 1-year monitoring required specific graphical and map processing to compare the information provided by POCIS and grab samples. Graphical projections demonstrated that with POCIS the number of quantified pesticides and the quantification frequencies were higher than with grab samples. Additionally, projections showed that POCIS provided better temporal representativeness of monthly contamination levels. Indeed, the POCIS data showed seasonal trends which were directly linked with the use of each pesticide (application period) and the land use of each sampling site, that was not visible with the grab samples data. Map projections of the measured concentrations, using a common scale for the two sampling strategies, clearly showed the strengths of the POCIS deployment and the link between measured contamination levels, quantified pesticides and land use. Finally, this study shows that the combination of grab sample data (magnitude of contamination peaks) and POCIS data (average concentration over a given period) provided more complete and reliable knowledge of the contamination levels in the Basin than either method alone.