Integrated assessment tools are used ever more frequently for solving new environmental, social, and economic challenges related to agriculture sustainability. This is particularly relevant in ecologically vulnerable areas, where mitigation options include a complete redefinition of farming systems. This paper presents an assessment of mitigation options of a coastal agricultural watershed affected by algal proliferations (Lieue de Greve, western France). We developed a methodology based on two existing assessment approaches, an agrarian system diagnosis and an environmental assessment, to identify connections between farming systems and risky nitrogen-agricultural practices and to explore potential ways to reduce the environmental impact of farms while respecting their economic viability and technical feasibility. To understand the functioning of farming systems and calculate key indicators of economic performance, the agrarian system diagnosis combines landscape, historical, and techno-economic aspects of a farm. The environmental assessment includes (i) calculation of farm-gate nitrogen (N) budgets and N-use efficiency and (ii) use of a spatially explicit biophysical model, which simulates the effect of agricultural practices on water and N dynamics at the watershed scale and on N fluxes at the outlet. At the farm scale, the main outputs of the approach were (i) the identification of 11 farming-system types and factors influencing them in the past and present, (ii) the assessment of their techno-economic performances and (iii) for the main dairy systems, the assessment of their potential environmental impacts. Insights about the capacity of some systems to adapt to new constraints linked with environmental objectives were also brought out. Grass-based systems appeared to be the best alternative for existing farming systems since they display good economic results while limiting N emission risks. Scenarios of changes in agricultural practices at the watershed scale were tested with the biophysical model, comparing them to a reference scenario based on the continuation of current agricultural practices. The main results indicated that: (i) current water quality was in equilibrium with current agricultural practices, (ii) the response time of the watershed to changes in agricultural practices was relatively short (6-7 years) and (iii) the expansion of an agro-environmental incentive based on low-input practices and grass-based fodder proposed by local stakeholders to the whole watershed would result in a significant decrease in N fluxes but would not fulfill water-quality objectives. This integrated assessment approach demonstrated its ability to promote the emergence of mitigation solutions and to improve decision support.
