A look at the DEPHY EXPE 2 projects presented.

The Carrefours de l'Innovation Agronomique (CIAg) organised a presentation of work carried out as part of the DEPHY Expé programme, resulting from system experiments conducted over a six-year period. These projects tackled a daunting question: "building agricultural production systems that use pesticides as a last resort". A daunting question, because behind the simplicity of the statement lies a series of difficulties linked to the history and practice of agriculture, issues of perception and even basic principles of ecology (agriculture being a very specific form of living system). This issue is not really a matter for research, industrial innovation, practical knowledge or public policy, but a little of all of these at once. It requires these different fields of action to meet and coordinate within a shared framework. It cannot be resolved in a top-down system in which research provides solutions that are then transferred to practitioners within a firmly established regulatory context, with the ultimate responsibility resting with the farmer, who is ultimately the only one to commit his salary. Before discussing how this issue has been dealt with by the various projects, it is useful to recall a few things. Firstly, research does not directly provide solutions; it first produces knowledge. Using knowledge to solve a problem is engineering and, as most of the presentations showed, engineering is a noble profession that deals with complex issues. Everyone knows the term civil engineering, and it would never occur to anyone to ask a researcher to build a bridge. It is high time that the term agro-ecological engineering (or agroecological engineering) takes its place to define the work that goes from the knowledge acquired to the practical implementation of transformations in crop protection. Secondly, the notion of transfer, present in all current projects, is also misleading. Take any response to a call for projects and read workpackage 3 or 4, often entitled "transfer and deployment", and you'll find the usual brochures, webinars, training courses and demonstration visits. That's all very well, but it's totally inadequate for tackling an issue as complex as the adoption of innovations with a view to redesigning crop protection systems. Supporting innovation remains a major difficulty in the transition to healthier agriculture. This is an issue that needs to be addressed both by research, with specialists in the social sciences, and by development and advisory services. Interesting initiatives exist, for example as part of the National Vineyard Depletion Plan (Les 15 du plan) or in certain networks set up by the Agricultural Technical Institutes. A good example was given during the seminar with the Cafés Agro (Cardona et al., 2024; Szklarek, 2021). Thirdly, we must not forget that the current regulatory framework was established during the era of chemical protection, from which we are being asked to emerge. A very strict precautionary principle is justified in the case of molecules that are by definition toxic to living organisms, but should it remain the same in the case of biological solutions? Why is it that a biocontrol product developed in three years under an emergency plan (the PNRI Betterave) has to wait eight years for approval (in this case, the system of derogations for experiments solves the problem, but without really securing the company's economic model)? Why not market vine plants protected by an attenuated strain of the dwarf drought virus ? What is the problem with using the hypovirulent strain of chestnut canker, which is naturally occurring ? We must not minimise the risks associated with biological approaches, which are real, but the re-design effort required of producers should be of the same intensity in all areas of action, including regulation.