Integrating agri-environmental indicators, ecosystem services assessment, life cycle assessment and yield gap analysis to assess the environmental sustainability of agriculture.

[Display omitted] • Different environmental assessment methods exist providing different viewpoints on the impacts of policies on the agroecosystem. • However, a general framework providing an holistic view is still missing. • Despite of their differences, existing assessment approaches can be combined to provide a more complete assessment. • We developed a conceptual framework describing the socio-ecosystem and the links between agriculture and the rest of the word. • We evaluated a set of environmental issues using the combination of the four existing assessment approaches demonstrating the interest of hybridizing the four existing approaches. Agriculture's primary function is the production of food, feed, fibre and fuel for the fast-growing world population. However, it also affects human health and ecosystem integrity. Policymakers make policies in order to avoid harmful impacts. How to assess such policies is a challenge. In this paper, we propose a conceptual framework to help evaluate the impacts of agricultural policies on the environment. Our framework represents the global system as four subsystems and their interactions. These four components are the cells of a 2 by 2 matrix [Agriculture, Rest of the word]; [Socio-eco system, Ecological system]. We then developed a set of indicators for environmental issues and positioned these issues in the framework. To assess these issues, we used four well-known existing approaches: Life Cycle Assessment, Ecosystem Services Analysis, Yield Gap Analysis and Agro-Environmental Indicators. Using these four approaches together provided a more holistic view of the impacts of a given policy on the system. We then applied our framework on existing cover crop policies using an extensive literature survey and analysing the different environmental issues mobilised by the four assessment approaches. This demonstration case shows that our framework may be of help for a full systemic assessment. Despite their differences (aims, scales, standardization, data requirements, etc.), it is possible and profitable to use the four approaches together. This is a significant step forward, though more work is needed to produce a genuinely operational tool. [ABSTRACT FROM AUTHOR]