Skip to Main content Skip to Navigation
New interface

Évaluation agroécologique des pratiques d’intégration agriculture-élevage: Application de l’analyse de réseau écologique à des exploitations agricoles en milieu tropical humide

Abstract : Agricultural systems have to produce more and better in a changing world. This requires adaptation to the increasing scarcity of natural resources while responding to the demands of a growing global population. In light of this, the design of local sustainable agricultural systems is a crucial task. Such agricultural systems should be simultaneously efficient, productive, resilient and self-sufficient, based on the principles of agroecology. This thesis presents an agroecological analysis of crop-livestock integration (CLI) in mixed crop-livestock farming systems (MCLS). We question to what extent, at farming system level, CLI, as a set of agricultural practices mobilizing ecological processes, can contribute to the agroecological performance of MCLS, and beyond, to agricultural issues as a whole. To answer this, we examine this question in the light of two hypotheses: i) CLI can be analyzed as a network of nutrient flows, with structure and functioning properties similar to those of ecosystems; ii) the most integrated MCLS, i.e. the most diversified and complex flow networks of nutrients, contribute to the agroecological performance of MCLS, enhancing productivity, efficiency, self-sufficiency and resilience. As part of our study, we first develop a methodological framework based on ecological network analysis (ENA) to assess the properties of agrosystems. We tested that framework on eight theoretical case studies, which represent a gradient of CLI and diversity of agricultural activities. We thus validate a grid of indicators to characterize both CLI and agroecological performances at system level. We then apply this methodological framework to seventeen MCLS in the humid tropics, from three territories with contrasting conditions in terms of access to production factors: Guadeloupe (F.W.I.), Brazilian Amazonia, and Cuba. The MCLS studied were modeled into nitrogen flow networks, on a year scale, rebuilt following iterative investigations with farmers. Based initially only on farms from Guadeloupe (eight farms), we show that the level of CLI is low at farm scale, although CLI practices appear significant at agricultural practice scale (animal feed from crop residues, and organic fertilizer from manure). We then extend our analysis to all seventeen farms from the three fields of study and show that CLI can be assessed on two criteria: the organization of flow networks and the intensity of flows circulating between agricultural activities. Our sample covers contrasting CLI situations, according to these two criteria. The trends that emerge from the analysis of the links between CLI and performance suggest that less integrated farms are inefficient and poorly resilient, or not very productive and resilient. However, highly integrated farms combine a high level of productivity and efficiency and an intermediate level of resilience. We demonstrate the benefits of ENA for the analysis of the functioning and agroecological performance of agricultural systems, even if such performance is not wholly dependent on CLI practices. In fact, the performance depends partly on the combination of agricultural activities and on the relative performance of these activities. Finally, simulation modeling of CLI practices within a given MCLS sheds greater light on the consequences of different forms of CLI on system performance.
Complete list of metadata
Contributor : Migration ProdInra Connect in order to contact the contributor
Submitted on : Friday, June 5, 2020 - 4:06:19 PM
Last modification on : Wednesday, March 23, 2022 - 12:08:32 PM


  • HAL Id : tel-02798473, version 1
  • PRODINRA : 355944


Fabien Stark. Évaluation agroécologique des pratiques d’intégration agriculture-élevage: Application de l’analyse de réseau écologique à des exploitations agricoles en milieu tropical humide. Sciences du Vivant [q-bio]. AgroParisTech, 2016. Français. ⟨NNT : ⟩. ⟨tel-02798473⟩



Record views