Characterization of agricultural spaces engaged in proximity-based alternative food networks: The southeastern France case study

Alternative Food Networks (AFN) are food commercial strategies based on geographical proximity. In Europe, these alternative strategies are emerging in a context of increasing productivity and the globalization of food markets. This article characterizes the forms of agricultural spaces involved in AFN within the PACA (Provence-Alpes-Côte d′Azur) region of France, and assesses the potential for agricultural spaces to engage in such alternative food chains. Based on interviews and semi-automatic data collection, the article develops a machine learning classification model to assess the likelihood of AFN engagement for agricultural spaces. The measurement of LISA (Local Indicators of Spatial Association) coefficients, based on the model's output, identifies differences between spatial clusters of conventional and alternative agricultural spaces. Alternative agricultural spaces are found to be more isolated than conventional compact regions. The model's analysis highlights the exclusion of alternative agricultural spaces from large production regions and their geographical proximity to urban areas. The results also underline smaller sizes, lower specialization, and more complex morphological characteristics for these alternative agricultural spaces, making them less suited to intensive forms of production. In light of strong public policy interest in food system relocalization, this article highlights the relationship between agricultural spaces characteristics and their engagement in AFN, and emphasizes the importance of considering the agricultural model when designing strategies to support and preserve 'local' agriculture.