Simulating nitrate loads and concentrations in the main French rivers by 2050 according to different territorial agricultural scenarios.

In excess compared to the other major nutrients, nitrogen (N) ­-transported mainly in the form of nitrates- from land surfaces to coastal areas, is responsible for harmful algal blooms in coastal areas. In France, most of the water loads of N transferred to the sea comes from fertilization on agricultural surfaces and their highly leached surpluses. The distance between the sources of nutrients and the manifestations of eutrophication is one of the difficulties encountered in reducing coastal eutrophication. The present study, which aims to model the N cascade at a national scale, for the current situation and according to different agricultural/dietary scenarios, is a basis for discussion with river-basin managers, who are required to enforce the Marine Strategy Framework Directive that sets the conditions for good status of marine environments.

Current N surpluses (2015-2018) are calculated via the GRAFS (Generalized Representation of the Agro-Food System), method (Le Noe et al. 2017), at the NUTS-2 scale (French regions) for arable land, permanent grassland and permanent crops. An approach based on yearly run-off coupled with a riparian denitrification routine allows linking these surpluses to average baseflow/quickflow concentrations in elementary catchments (average size: 5.8 km²). These diffuse lateral inputs, plus point-source inputs, are then integrated into the pyNuts-Riverstrahler modelling platform (Billen et al. 1994, Raimonet et al. 2019), which calculates nitrogen transformations and transfers along the river network. The model outputs have a kilometric spatial resolution and a 10 days temporal resolution, making it possible to finely represent the spatio-temporal variability of N concentrations and loads on the French territory.

Different plausible scenarios for the future of the agro-food system in France, with a gradient of territorial specialisation/N intensities, are then modelled within the GRAFS-pyNuts-Riverstrahler approach to assess their impacts on the N cascade at the national scale. This study is part of the Nuts-STeauRY project (funded by the OFB: https://nuts-steaury.cnrs.fr/), which aims to demonstrate the contribution of integrated land-sea modelling to the realisation of spatialized scenarios to limit coastal eutrophication in France. Carbon, phosphorus and dissolved silica will also be modelled for this project.

Billen2014:10.1007/BF00007414

LeNoë2017:10.1016/j.scitotenv.2017.02.040

Raimonet2019:10.3389/fmars.2018.00136