A protocol for the conceptualisation of an agro-ecosystem to guide data acquisition and analysis and expert knowledge integration

Abstract

Innovative agricultural systems need to combine the production of goods with the provision of environmental services. When agronomists analyse or design multifunctional agro-ecosystems, they thus need to include knowledge of an increasing range of scientific disciplines (plant biology, soil science, ecology, etc.) while continuing to use their systemic approach as a cornerstone. Increasing amounts of knowledge of different types (concepts and data) will thus have to be included in systemic approaches that are developed in the agronomic domain. Knowledge integration and sharing are frequently hampered by the lack of detail in the assumptions made in each discipline. We hypothesise that a standardised description of the conceptual model underlying data collection and the analysis of agro ecosystems would improve transparency and knowledge integration. Here we propose a protocol to formalise the conceptual modelling of an agro-ecosystem (CMA) related to a specific agronomic issue. The CMA protocol is implemented in four iterative steps: (i) structural analysis, (ii) functional analysis, (iii) dynamic analysis, and (iv) consistency check. The final product is a conceptual model of an agro-ecosystem whose key elements are a structured knowledge base and associated graphical representations. The protocol was drawn up based on three case studies concerning three different biophysical objects (coffee agroforest, cotton, grapevine) with different problems to be addressed. They are given here as an illustration of how to apply the CMA protocol, and to show how it can be used as a tool to build a systemic representation of a complex agro-ecosystem, as a tool for agronomic diagnosis and yield gap analysis, or as a tool to elicit a range of expert knowledge to design new field experiments. The CMA protocol proved to be efficient in guiding the process of conceptualisation up to the point at which the variables that need to be measured in the field are identified and interlinked. It enabled elicitation and integration of knowledge from different biophysical disciplines and different types of expertise during the conceptualisation process. It also enabled identification of knowledge gaps, and the design and analysis of experiments to tackle complex problems. The CMA yielded by the protocol could be used again, thanks to its transparency and modularity. Further work is underway to improve the CMA representation and its uses in numerical model specification and in participatory methods for the design of cropping systems.