Thibault SALOU will defend his PhD in front of the following committee:
- Reviewer: Enrico BENETTO (LIST)
- Reviewer: Arnaud REYNAUD (INRA)
- Member: Claudine BASSET-MENS (CIRAD)
- Member: Philippe LE GOFFE (AgroCampus Ouest)
- Member: Thomas NEMECEK (Agroscope)
- Member invited: Vincent COLOMB (ADEME)
- Supervisor: Chantal LE MOUËL (INRA)
- Co-Supervisor: Hayo VAN DER WERF (INRA)
Abstract: Life Cycle Assessment (LCA) is a normalised and multicriteria method that assesses environmental impacts of goods and services. It was first used for benchmarking, environmental communication and support for product development within the framework of Attributional LCA (ALCA) which assesses impacts in a status-quo situation. In the early 2000s, methodological development led to a new type of LCA, Consequential LCA (CLCA), which assesses environmental impacts of changes in the system under study. CLCA has been used to assess impacts of policy changes, especially in the energy and biofuel sectors. Indeed, by taking market effects into account, CLCA is an appropriate tool to quantify direct and indirect impacts induced by policy changes. However, although CLCA principles are now well accepted by the LCA community, methodological debates about its practical implementation remain, especially for the definition of system boundaries and identification of affected markets.
This Ph.D. thesis combines LCA and economic modelling to assess environmental impacts of policy instruments in the dairy and wider ruminant sector in France and the rest of the European Union (EU) using CLCA. We focus on dairy production because it is one of the main contributors to environmental impacts, such as climate change or eutrophication, in the EU. Moreover, policy instruments that concern this sector are changing due to recent Common Agricultural Policy reforms. These reforms aim to promote production sectors that have both more market orientation and less environmental impact.
First, we identified seven dairy production systems that represent French production and estimated their environmental impacts using ALCA. Results revealed that their environmental impacts varied greatly depending upon the functional unit considered. Consequently, for accurate assessment of multifunctional systems, such as agricultural systems, we recommend using both mass-based and area-based functional units.
We then modelled the seven dairy systems in MATSIM-LUCA, a partial equilibrium model representing agricultural markets at the global scale, to estimate effects of removing EU dairy quotas on the shares of different types of dairy systems in French milk production. We also assessed sensitivity of these results to different contexts of global demand for dairy and meat products. The main finding was that changes in production context had little influence on proportions of dairy system types in French milk production. The systems do not react equally, however; intensive maize-based dairy production is the most sensitive to changes in production context.
Next, we used MATSIM-LUCA to assess market effects of two changes in EU policy instruments: removal of dairy quotas and establishment of a grassland premium. Model predictions (i.e. changes in yields, types of production, surface areas of production, trade, animal diet composition) were used as inputs to a CLCA model. We assessed effects of changes in policy instruments on a range of environmental impacts (climate change, eutrophication, etc.). We also demonstrated the value of using economic modelling - especially MATSIM-LUCA, because of its production technologies specifications - to support CLCA. In fact, the model can identify not only production displacements and substitutions but also changes in production technologies (intensification, changes in animal diet composition, etc.). It is important to consider all of these effects, as they can influence environmental impacts of goods but also dynamics of land-use change.
Finally, we identified improvements that are needed in LCA methodology and MATSIM-LUCA. For LCA, improvements include defining one or more functional units to assess multifunctional systems and developing methods to bridge data gaps to perform CLCA. For MATSIM-LUCA, improvements include strengthening links between LCA and economic models by representing the supply of production factors in more detail, and improving representation of land markets for more accurate assessment of land-use changes.