2025 Volume 61 Issue 1 Pages 49-55
Whether considered as a science, a set of agricultural practices or a social movement, agroecology calls for new ways of doing research and training. Drawing on key elements of the literature, this article sets out to illustrate the ‘agroecological turn’ in research and training on agricultural and food systems. Two examples of the co-creation of knowledge and innovations, in interaction with training, are presented and discussed: a European project to increase the use of agrobiodiversity in farms and food value chains, and the French network of experts on short food supply chains, bringing together research, development and training organizations. The first example illustrates a systemic and participatory approach from seed to plate, anchored in multi-stakeholder living labs, in which agroecology is a catalyst for the collective redesign of food systems. The second example shows how the national network supports and up-scales local initiatives around short food supply chains conducive to the agroecological transition, including by identifying training gaps. These two examples open up a discussion on two issues raised by place-based research on agroecology, namely stakeholder participation and the articulation of scales to transform food systems.
Depending on the period, country and discipline, agroecology has been seen as a science, a social movement or a set of agricultural practices (Wezel et al., 2009). According to Gliessman, a major figure in this field, these three approaches are rather three dimensions of agroecology, to be integrated to understand the transformation of food systems (Gliessman, 2018). Moreover, beyond the debates on its definition, debates on agroecology also concern the way in which research and training are, or should be done, in this area. The aim of this article is to contribute to these debates1, which are part of what we propose to call the ‘agroecological turn’ in research and training on agricultural and food systems.
In the first part of this article, we present inputs from the literature on the changes of approaches for research into agroecology, stressing the importance of participatory approaches at local level. In the second part, we develop two cases of co-creation of knowledge and innovation on agroecology, a European project on agrobiodiversity, and the French national expert network on short food supply chains. In the third part, we discuss the challenges about stakeholder participation, but also the importance of linking different scales to transform agricultural and food systems.
Agroecology is the subject of a vast scientific literature. The aim here is not to review the state of the art, but to highlight some of the salient aspects of literature reviews on the new avenues of research and training opened up by agroecology. Reviewing a large number of studies, Meynard (2017) highlights four dimensions that characterize the change required by agroecology in the conduct of research and training. First, agroecology involves a systems approach, taking into account the interactions between the components of systems. Second, it calls for local empirical knowledge to be valued, particularly in relation to the observation of ecological processes. Third, it requires agriculture and farms to be seen in the context of food systems, because they are not isolated entities; agriculture and farms are influenced by the organizations they are part of, and vice versa. Fourth, agroecology encourages collective learning, to make the most of everyone’s specific knowledge (Meynard, 2017). In addition, the first review of scientific studies on farms transition to agroecology brings out innovative research paths, in which farmers’ creativity and experiments are better valued (Prost et al., 2023).
(2) A focus on participatory research at local levelThe systems perspective required by agroecology can be developed through a diversity of methodologies, ranging from participatory tools and field observations to mathematical simulation modelling (Tittonell, 2023). However, participatory research at local level has been the subject of a growing body of work in the last decades. Lopez-Garcia et al. (2021) emphasize that the transformative standpoint of agroecology calls for science ‘to be done with people’. Research into agroecology thus contributed to strengthen the development of living labs (McPhee et al., 2021), ‘in which experimentation is conducted on real farms, in specific territorial and community contexts, with farmers and other actors involved from the beginning as equal partners in proposing ideas, testing them, improving them and promoting them further’ (European Commission, 2024). However, stakeholder participation in research on agroecology remains challenging (López-Garcia et al., 2021), while the place-based co-creation process of knowledge and innovations in living labs questions their ability to transform food systems at larger scale (Massari et al., 2023). Using two examples from France and Europe, we propose to illustrate the changes in approach highlighted in the literature and to provide some answers to the issues raised by the participatory and local dimensions of research and training on agroecology.
The first example is the European DIVINFOOD project, aiming to co-construct short and mid-tier food chains that value varieties and landraces cultivated with agroecological practices and contribute to sustainable food systems. Developing agroecology requires biological diversity on farms and in the countryside (Altieri et al., 2015). However, the agro-industrial model has led to the decline of wild but also cultivated biodiversity (IPBES, 2019), with just three plant species—wheat, maize and rice—providing 60% of the plant calories we consume worldwide (FAO, 2012).
The project aims to produce knowledge and innovations to increase the use of neglected and underused crops (Padulosi et al., 2021), in this case minor cereals and legumes, from seed to the consumer’s plate. These two families of species have a great potential for the agroecological and food transition (Ditzler et al., 2021; Gazza and Nocente, 2023). However, there are many scientific and technical obstacles: first, farmers lack of varieties as these neglected crops did not receive much attention from breeders. Moreover, the varieties available have been selected in conventional agriculture and are often little adapted to agroecology. Secondly, there are only a few agronomic references for neglected and underused species and no specific guidelines for processing and marketing these little-known crops. All the stages in the value chain need to be studied, but, in line with Meynard’s approach, not separately: farmers produce in order to sell, and value chains need to integrate the expectations of consumers—who are also citizens with moral values—right from the selection of varieties. Then, the DIVINFOOD project started with a large consultation at the European level to capture citizen-consumers’ expectations, and aversions, regarding the use of cultivated biodiversity in food chains (Chiffoleau et al., 2024).
To develop this systems approach, the project is bringing together all skills in multi-stakeholder living labs, which enable to test varieties, farming systems, processing techniques, marketing modes and education tools under real conditions, to capitalize on existing initiatives and knowledge, especially farmers’ own experiments, and to co-develop technical, technological and organizational innovations. In these living labs, research is not a substitute for local knowledge, nor does it take people’s place. Knowledge and innovations adapted to local situations are created through the meeting between empirical and scientific knowledge (Darré, 1994). Participation is not limited to professionals of the agri-food sector: citizen-consumers also take part in the activities of living labs (participatory breeding, tasting of new food products, etc.), as well as local authorities, but also medical doctors, artists, etc.
Collaboration between researchers and stakeholders is producing knowledge about varieties and cropping systems suited to agroecology (intercropping, agroforestry, etc.); about minimal processing methods that respect the inherent quality of the crops (e.g., fermentation of legumes); and about the marketing methods and consumer education needed to enhance the value of agroecologically produced products. However, knowledge production follows a loop, in which, for instance, discussions with sellers and citizen-consumers (i.e., the end of the supply chain) guide the plant breeding work (i.e., the beginning of the supply chain). It is a loop, but also a crossroads between three domains of research and innovation, and a meeting between all concerned actors, which collective training sessions help to bring about. In these sessions, farmers, processors, advisors and researchers learn from each other what they are doing and how, which, notably, facilitates ‘coupled innovation’, both in agriculture and in processing (Meynard et al., 2017). For example, new links have been developed between farmers and a Japanese chef living in France, who was able to make dorayakis (traditional Japanese cakes with bean paste) with ‘meat bean’, a bean that had almost disappeared in France; a bean which the farmers involved in the DIVINFOOD project are now growing again.
However, new business models have also to be developed to both value small-scale agroecological production and keep affordable prices for lower budget consumers, to prevent agroecology from becoming an elitist and exclusive field. Three approaches have been identified in DIVINFOOD: cooperation between farmers to reduce production costs (e.g., by sharing sorting equipment), support by the local authorities (e.g., through payment for ecosystem services), fair partnerships with buyers, among which restaurants and collective catering. The latter are strengthening the business models by also developing new, attractive and tasty recipes with legumes, whose acceptance by consumers remains low in Europe (Henn et al., 2022). Agroecology is much more than a fusion between agronomy and ecology, it is a catalyst of the redesign of food systems.
(2) The French expert network on short food supply chainsThe second example is the national network of experts in short food supply chains, launched in France in 2015. This network is funded by the French Ministry of Agriculture and Food Sovereignty, and co-led by Civam (a network of alternative farmers) and INRAE (research organization). It brings together research, training and development players involved in analysis, support or training on the theme of short food supply chains developed at the local level, questioning both the diversity of these chains and their contribution to sustainable food systems.
Short food chains have long been marginalized in France as they were associated with archaic forms that rejected technological progress, or with forms of protest (Chiffoleau, 2019). Yet these sales methods are part of the history of the country and of its agriculture, and are closely linked to agroecology. In France, where short food chains are used by ¼ of farms (i.e., about 100 000 farms in 2020), there are 4 times more organic farms in these channels than in long supply chains (Barry, 2023).
Short food chains will not replace long ones. Above all, they do provide key opportunities for systems research into the social and economic mechanisms involved in the agroecological transition. When farmers use short channels, they diversify their sources of income and reduce the financial risks, that favors risk-taking in farming practices. Moreover, farmers have the opportunities to talk to consumers who are concerned about their health, and who recognize the efforts of farmers to ‘produce better’ while exerting a soft pressure on them to maintain good practices. Finally, short food chains are also important opportunities for discussion between peers and for learning from the farmers who are the most advanced in agroecology. All these mechanisms are conducive to the agroecological transition (Chiffoleau, 2019).
However, experts involved in the national network have identified training gaps, whether for future and current farmers, student engineers or technicians who will become advisers, or processors who need to learn how to manage agroecological raw materials. Initial training courses in France have not yet integrated the issues related to short food supply chains. Farmers are generally not trained to combine several sales channels, and to manage agricultural diversification in order to propose a wide range of products to consumers. Training courses do not prepare agricultural or food technology engineers and technicians to advise on setting up these channels in agroecological conditions. In this context, the raw material is necessarily heterogeneous, and the solution is not to use synthetic inputs in fields or additives in processing to reduce this heterogeneity, but rather to value the inherent quality of heterogeneous crops through minimal processing (de Vries et al., 2018). The network values and facilitates the organization of multi-actor training sessions (e.g., mixing breeders and butchers), in which the stakeholders understand the work and the working conditions of the others. Last but not least, these collective learning sessions also provide opportunities to collect local knowledge about the implementation of agroecology in real settings, and thus feed research.
These two cases illustrate the changes in approach highlighted in the literature, but also raise points for discussion.
Stakeholder participation remains a voluntary act. The DIVINFOOD project is structured into 9 living labs and the participation of farmers, processors, consumers and local authorities varies from one living lab to another. The relationships of the project members with the professional and local communities play a major role in the ability to mobilise a large number and range of people. We also observe that organic farmers and agricultural organisations have difficulty exchanging with non-organic farmers and organisations within the local food systems, as highlighted in other contexts (Coq-Huelva et al., 2017). Even if agroecology cannot be reduced to organic farming, the latter is a key driver for the agroecological transition, but the lack of relations between organic and non-organic players limits its impact.
More generally, participatory research faces a twofold challenge: mobilising and retaining participants, but also enabling participants to play a real part in the activities and associated decisions (on research hypotheses, study protocol, etc.). Agroecology raises specific issues with regard to these challenges, precisely because it can be equated with organic farming, thus limiting the participation of those who are not organic. Moreover, this subject concerns everyone but raises highly technical questions, so that citizens often do not have the knowledge to implement ‘critical participation’. Based on sociology of organisations, critical participation is when actors do not simply accept decisions that have already been taken by others, but are also aware of the issues underlying the activities implemented, and are able to make proposals (Friedberg, 1988). In the DIVINFOOD project, knowledge input is provided to citizen-consumers prior to any consultation, and this remains a point of vigilance.
In terms of mobilising and maintaining participation, the actor-network theory provides concrete resources to develop and stabilise a multi-actor network around agroecology, following several steps (Callon, 1984): sharing the common problem/challenge (e.g., decline in agrobiodiversity), getting each type of stakeholder interested by translating the interest for the collective into interest for each one (e.g., contribution of legumes to soil fertility and healthy diets), and enrolling individuals by valuing their specific and unique contribution: research in agroecology prompts to recognise farmers as practical experimentalists of nature-based solutions (MacMillan and Benton, 2014), while the contribution of consumers co-creating knowledge and innovation in short food chains can be explored and valued in the prosumption perspective (Podda et al., 2021).
(2) Articulating scales to transform food systemsA second line of discussion is about the limitations of place-based co-creation of knowledge and innovations. On one hand, person-to-person interaction at local level enables learning and co-creation, as well as the exercise of soft social control and emulation that regulates and stimulates collective innovation (Lazega, 2001). On the other hand, the development of agroecology comes up against technical, economic and regulatory lock-ins maintained by the upper-scale regime supporting the agro-industrial model (Vanloqueren and Baret, 2009). The national network on short food chains plays a key role in formalizing and passing on to policy-makers the proposals arising from bottom-up local approaches. For example, a major breakthrough has been achieved on the regulatory front, to adapt regulations for on-farm processing of animal products to real-life conditions, and no longer to agro-industrial processing standards. This change in regulation is not so much linked to “pressure” exerted on the regime by innovation niches, as put forward in the multi-level perspective (Geels and Schot, 2007), but is rather the result of in-depth documentation of the real, local conditions under which on-farm processing is practiced. This documentation involved advisors and researchers within the national network supported by institutions participating in the regime. Niches and regimes don’t operate in isolation, they interpenetrate, and participatory research on agroecology can play a key role in strengthening this interaction.
With regard to another lock-in, the lack of recognition of agroecological products on the market, the national network also provided an opportunity to exchange local experiences, such as the Ici.C.Local approach, which consists of signaling, on open-air markets and by a color code, the local products coming from short food chains and respecting principles of agroecology (Chiffoleau et al., 2016). Based on local management and a participatory guarantee system, the approach is spreading in France and has been highlighted by the FAO as an interesting tool for developing agroecological markets (FAO and INRA, 2018).
These two examples demonstrate the importance of combining different scales to transform food systems. Local initiatives are appropriate scales to co-create knowledge and innovation in agroecology, while the national network facilitates its scaling up, combining deep scaling (e.g., changes in regulations) and out scaling (e.g., Ici.C.Local propagation) (Moore et al., 2015). However, this calls for more research into the conditions of articulation between these scales, between new power relations and compromises.
Agroecology is gaining interest worldwide as a response for food systems to meet global challenges (FAO, 2018). Agroecology is also shaking up linear, top-down approaches to research and training, by promoting local knowledge and co-creation. In this article, based on highlights from the literature, we illustrate these ways of doing research, in interaction with training, but we also raise a number of issues that need to be discussed and explored in greater depth. The agroecological turn in research and training on agricultural and food systems is ultimately as much an empirical dynamic as it is a heuristic approach to questioning and renewing our research approaches and agendas (Tittonell, 2023).
The DIVINFOOD project is funded by the Horizon 2020 research and innovation program of the European Union under grant agreement No. 101000383. The French national expert network on short food supply chains is funded by the Ministry of Agriculture and Food Sovereignty, as part of the Special Allocation Account for Agricultural and Rural Development.
