Abstract :
[en] In a world facing a growing population and diminishing resources, agriculture has a major role to play in terms of sustainable provision of food and feed, job creation in rural areas, preservation of natural resources and climate change mitigation. As a very intensive agricultural region, Wallonia (southern Belgium) dedicates more than half of its arable land to cereal crops. Thanks to a long history of research and development, rich soils and favorable climate conditions, yields of cereals such as wheat or barley cropped in Wallonia are amongst the highest in the world. As a major sector in the Walloon agricultural landscape, the cereal sector therefore calls for production and processing opportunities that meet the requirements for sustainable agriculture.
In this context, the objectives of the present thesis were to evaluate the environmental sustainability of current and potential evolution scenarios for the Walloon cereal sector using the Life Cycle Assessment (LCA) methodology fed with local data.
On the basis of current uses of cereals grown in Wallonia, this research aimed first at describing various scenarios for the future of the Walloon cereal sector. Based on contrasting hypotheses, four scenarios for the uses of Walloon cereals by 2030 were defined with the support of stakeholders involved in all parts of the sector. Issues such as competition between food uses (both direct and indirect) and non-food uses or Walloon dependency on cereal imports were addressed in these scenarios, which illustrate contrasting future situations, from food self-sufficiency to drastic globalization, through the development of new outlets combining changes in crop-growing practices and consumption habits and new technologies.
Using the widely used, yet continuously improved, methodology of LCA, the second part of this work evaluated the environmental impacts of cereal production in Wallonia. It identified the production steps with the greatest contribution to environmental impacts. The leading contributor was mineral fertilizer production, for which this work used up-to-date data from best available technologies and demonstrated the importance of using such recent data. The second greatest contributor to environmental impacts was emissions from mineral and organic fertilizer use on field, before and during plant growth until harvest. As well as being very important, these emissions are also extremely variable, depending on agricultural practices, soil and weather conditions, and thus particularly difficult to assess. This is particularly true of nitrogen- and carbon-related emissions, whose dynamics are very problematic to evaluate. This work demonstrated the limits of commonly used emission models, which use generic emission factors and therefore weaken LCA results. As a result, the use of more specific models, based on field trials and adapted to local conditions, was placed at the top of the list for future research in this area. From a methodological point of view, this part of the work also confirmed the influence of the functional unit used to express LCA results. It showed the strong link between yield and environmental impacts, and identified cereal crops demonstrating efficient input management, with lower impacts per kilogram of product, per hectare and per euro. Other methodological choices, such as the choice of allocation factor between grain and straw, also showed their relevance. In short, this part of the work demonstrated the pertinence of using local data to feed LCAs of agricultural productions and identified the most sensitive parameters to be adjusted in priority to achieve good quality LCAs.
Thirdly, on the basis of an original methodology comparing equivalent balanced animal diets, a case study on cereal processing in a biogas plant assessed the consequences of diverting feedstock from the animal feed industry into the bioenergy sector. Despite the environmental benefits brought by fossil fuel displacement and digestate use as organic fertilizer, the replacement of raw materials previously used as animal feed induces additional impacts potentially more important than the benefits of this technology. The key message from this finding is that it is urgent to stop using potentially edible raw material for bioenergy production. The only countervailing consideration here is that current trends towards decreasing meat consumption would in the longer term make lands and raw material available for purposes other than animal feed.
The strength of all the LCA results in this work was tested using a wide range of sensitivity and uncertainty analyses. These good practices, often neglected due to time constraints, proved to be essential to be able to draw robust conclusions.
Through the application of LCA to the Walloon cereal sector, this work had some practical implications for enhancing the quality of LCAs of agricultural productions. It showed the sensitivity of the method to the use of local data and demonstrated the strengths and limitations of the method. It has produced specific LCA references for cereal production in Wallonia, which are useful for feeding environmental impact assessments of cereal-based products sourced with Walloon cereal crops.
Further improvements could include the use of specific models to assess on-field emissions from fertilizer use and the consideration of additional impact indicators regarding soil organic stock changes, biodiversity impacts and water consumption. Additionally, assessing impacts at the rotation scale would reflect more accurately the way farmers optimize their crops and allow for the evaluation of more diverse cropping systems.
