Use of environmental management data for mass flow estimations of plastic debris in rivers: The Seine River and The Huveaune River, France.

Methods to quantify plastic transport in rivers have greatly improved during the past few years. As a first approach, visual counting is currently the simplest way to assess plastic transport with minimal effort and cost. It usually results in underestimations of plastic input into the sea of about one to two order of magnitude when compared to models such as the Jambeck’s approach. The latter shows statistical weaknesses and data availability issues leading to large uncertainties, while visual counting miss the water column compartment and often has a low spatiotemporal representativeness. In order to give another ground-truth estimation of plastic transport able to challenge both models and visual counting, we developed innovative methods based on environmental management data in the Seine estuary (500 m³/s) and the Huveaune River ( 2 m³/s; Marseille, France). First, we used data from institutional cleaning in the Seine estuary that consist in litter collection on riverbanks. Their efficiency was measured based on capture-recapture design. Mass flows of plastic debris were then calculated based on the capture rate over one year, the estimation of the fraction of plastic debris which are never collected (hidden or too small) and the assumption that all plastic debris strand on riverbanks. Second, we used data from bar screens spaced of 3 cm in the Huveaune, a small urban river flowing in Marseille, South France. All the water column is screened, and captured waste are automatically collected in dumpsters. Grab sampling were performed after a dry, a wet and a flood period. The corresponding annual mass flows of plastic debris was then calculated relative to the mean fraction of time corresponding to those hydrological periods over 2017 and 2018. Annual mass flows of plastic debris were normalized to the population in both basins. Although methods were different, mass flows of plastic debris per capita are very similar with 8.5 – 13.6 g/cap/yr for the Seine River and 2.4 – 14.9 g/cap/yr for the Huveaune River. This is one to two order of magnitude lower than the Jambeck’s approach. However, when focusing on the fraction ending into the Sea, bar screens in Marseille enable to decrease the mass flow of plastic debris of about one additional order of magnitude, while cleaning of riverbanks decreases it of about 10%. This is related to the nature of the rivers that calls for different solutions, screening the whole Seine River being a tricky idea. Nevertheless, when normalized to water volume, the Huveaune River is visually much more polluted (16.4–102.2 mg/m³) than the Seine estuary (9.0–14.5 mg/m³). In conclusion, environmental management data can help to estimate mass flows of plastic debris and calls for better consideration. However, they often need an improved scientific framework.