Research

INTRODUCTION

Transformation of agri-food systems is urgently needed to face current global crises (Mier y Terán Giménez Cacho et al. 2018, Egmose et al. 2021, Giller et al. 2021). Agroecology is seen by many as a key element of such a transformation, intertwining science, social movements, and agricultural practices (International Forum for Agroecology 2015, Food and Agriculture Organization (FAO) 2018, Wezel et al. 2020). Méndez et al. (2017) describe agroecology as “an approach that seeks to integrate ecological science with other academic disciplines and knowledge systems to guide research and actions towards the sustainable transformation of our current agrifood system.” To support the development of agroecological initiatives, transdisciplinary knowledge development thus has an important role to play (Levidow et al. 2014, López-García et al. 2021).

As a temporary collaboration of scientists and non-scientists with the goal of facilitating learning and promoting structural change, a reflexive arrangement aims to co-create new knowledge that is translated into joint transformative action (Hendriks and Grin 2007). Reflexive arrangements are situated in the specific context of a societal initiative (Guzmán et al. 2013, Méndez et al. 2016, Rossing et al. 2021) and recognize the importance of actor networks for change (Elzen et al. 2012, Magrini et al. 2019). A shared understanding of an initiative’s transformative history and current state provides valuable insights for joint action and contributes to building trust and social capital necessary for effective science–society interactions (Hoffecker 2021, Koole 2022).

In Europe, agroecology is part of the Farm to Fork policy of the European Commission (European Commission 2020). A range of policy instruments are in place to support transformations of the agri-food system by benefiting from agroecological principles. Among these, the Horizon 2020 research and innovation funding scheme and its successor Horizon Europe feature multi-actor, multi-country approaches through which researchers connect to societal innovators in what are essentially reflexive arrangements. Such transdisciplinary approaches are also sought after by global research donors, such as the Bill and Melinda Gates Foundation (BMGF) and the CGIAR (formerly the Consultative Group for International Agricultural Research). Analyses of the efficacy of research projects in bringing about societal change have shown a range of disabling factors at different levels (see, e.g., Cronin et al. 2022 for Horizon 2020 projects; Schurman 2018 for BMGF projects; Leeuwis et al. 2018 for CGIAR projects). A common shortcoming is a lack of an agri-food systems perspective as a basis for a theory of change that is legitimate for all involved in a reflexive arrangement (Zurek et al. 2023). With the aim of filling this gap, we present a heuristic, learning-oriented framework for systemically portraying societal agroecological initiatives at the start of reflexive arrangements. The framework was developed and tested in a Horizon Europe multi-actor project. We adopted the verb “portray” to emphasize the intention to combine scientific craftsmanship with local actor knowledge and objectives to arrive at a shared representation of the history and current state of the agroecological initiative. In developing the framework, we integrated both quantitative (e.g., López-Ridaura et al. 2002, Mottet et al. 2020) and qualitative approaches (e.g., Vanwindekens et al. 2013, Morel et al. 2020, Holmén et al. 2022, Rocker et al. 2022) for systems characterization and diagnosis. We also drew on participatory reflexive approaches (e.g. Engel 1997, Bos et al. 2009, van Mierlo et al. 2010, Elzen and Bos 2019, Rossing et al. 2021, Leclère et al. 2024) and approaches that address the evolution of societal initiatives (e.g., Gremler 2004, Britt and Wilson-Grau 2012, Coupaye 2015, Douthwaite and Hoffecker 2017, Polge and Pagès 2022).

Funded for 4 yr from September 2022 under the Horizon Europe funding scheme, the Agroecology-TRANSECT project connected researchers and 11 agroecological initiatives under the common objective of unfolding agroecology for Europe (https://www.agroecology-transect.net/). Each initiative consists of a network of societal and academic actors collaborating to develop systemic, agroecological solutions that address climate change mitigation, enhance biodiversity, and improve socio-economic resilience (i.e., the three overarching themes of the project). As selection criteria, the agroecological initiatives were chosen to cover a diversity of European geographical regions and farming systems. At the start of the project, each initiative had been functioning for at least 4 yr, showing the ability to function autonomously in terms of human and material resources. This selection process led to 11 initiatives representing a diversity of transformation levels toward agroecology, from efficiency gains to food system redesign (Gliessman 2016). The project was designed to engage participant scientists and those agroecological initiatives in learning cycles toward greater achievement of agroecological principles in real-world settings. The Danish initiative was selected as exemplary of the reflexive arrangements in the Agroecology-TRANSECT project. This initiative is a conservation agriculture (CA) network evolved because of farmer decision making motivated by keeping the yield potential while increasing economic net returns from reduced labor expenses (time in the field) and machine investment costs (Hansen et al. 2020). Using CA principles: (1) minimum mechanical soil disturbance by using reduced or no tillage; (2) permanent soil cover using cover crops and/or crop residues; and (3) spatial and temporal crop diversification through a variety of cropping strategies and crop rotations (FAO 2022), the network expects to meet political targets like 55–65% reductions in total greenhouse gas emissions from forestry and agriculture by 2030 (Ministry of Food, Agriculture and Fisheries 2021) and nitrogen leaching as part of ongoing Danish implementation of the EU water framework and nitrate directives (Environment Agency 2023). Through its collaborative and knowledge-sharing approach, the network is identified as a key driver for change across the sector, facilitating wider adoption of agroecological principles such as infield biodiversity enhancement, soil health improvement, and input reductions.

The objective of this study was to develop a heuristic framework to portray agroecological initiatives as a starting point for reflexive arrangements. In developing the framework, we aimed to balance thoroughness and ease of application to enable adequate time for subsequent project activities to benefit from its outcomes. The heuristic framework aimed to answer the following questions: (1) What is the context the initiative is embedded in? (2) Which actors are related to the initiative, how are they related to each other, and who are the key actors? (3) What are the barriers and levers for the development of the initiative, and which ones are key? We illustrate the use of the heuristic framework by portraying the Danish reflexive arrangement and discussing the way in which a shared understanding of the initiative’s transformative history and current state provided insights into options for joint action within the Agroecology-TRANSECT project’s mandate.

METHODS

Heuristic framework to portray agroecological initiatives for reflexive arrangements

Based on a systematic review of approaches used to evaluate projects and societal initiatives that aimed at transformative change, we propose a heuristic framework to portray agroecological initiatives in reflexive arrangements comprising three pillars: context, actors, and barriers and levers. Below we review each of these.

Context

Context reflects the situated nature of agroecological initiatives (Méndez et al. 2017, Barrios et al. 2020). Drawing on a variety of frameworks capturing socio-technical (Geels and Schot 2007, Ghosh et al. 2022), socio-environmental (Millenium Ecosystem Assessment 2005), food (Nesheim et al. 2015, shiftN 2023) and farming systems (Schoonhoven and Runhaar 2018, Escobar et al. 2019, Agroecology Europe 2020, Mottet et al. 2020), the context of agroecological initiatives is described by six broad dimensions, covering the biophysical environment, knowledge, society, policy and governance, economy and farming systems. Each dimension is subdivided into elements to specify relevant features of agroecological initiatives. In Table 1, the elements and key references are summarized.

Actor network

Social networks mediate agroecological transitions by acting as conduits of information, collaboration, and material resources (see review in Anderson et al. 2019a). The capacity of a social network to support innovation is determined by its structure and by the position of individual actors (Gaitán-Cremaschi et al. 2022). Studying relations between actors gives insight into strengths and weaknesses of the network and highlights possible levers of change through reorientation of relationships (Rocker et al. 2022).

Network of barriers and levers

Innovation scholars consider barriers as innovation system failures, slowing down system change and blocking actors in their learning (van Mierlo et al. 2013). A leverage points perspective reveals areas in complex systems for transformative change interventions (Meadows 1999, Fischer and Riechers 2019). Drawing from these definitions, we have defined barriers and levers as factors that negatively or positively, respectively, influence the implementation, operation, maintenance, scaling, or replication of an agroecological initiative. Barriers and levers for the development of agroecological initiatives can arise within an initiative or external to it (Schoonhoven and Runhaar 2018). Barriers often present potential levers for change. The identification of barriers and levers allows reflection on factors and their role in the development of the agroecological initiative and provides learning across initiatives (Holmén et al. 2022). Recognizing that barriers and levers are not mutually exclusive and may influence each other (Hurley et al. 2023), we approach them from a systems perspective, considering their connections over the course of time.

Case study: the Danish Conservation Agriculture Network

Since 2016, on the island of Zealand (Denmark), a voluntary demand-driven network of about 50 farmers with cereal-dominated crop rotations has been involved in developing CA on their farms. Their ultimate objective is to foster the adoption and expansion of CA by stimulating peer-to-peer knowledge exchange focusing on modified technical solutions. The farmers reduced soil tillage by not ploughing and keeping the soil covered after harvesting the main crop during autumn and winter. Rotational crop diversification is less developed due to animal feed dominating land-use traditions and connected markets (Hansen et al. 2020). Over the years, the farmers noticed a range of positive changes in the CA fields compared with traditional and neighboring tillage-based systems and became highly convinced of its benefits for their farm operations and for society at large (Hansen et al. 2020). At the same time, financial and social rewards remained low to absent, prompting questions about what factors blocked wider adoption. In Denmark, reduced tillage is practiced on 25.5% of the utilized agricultural area, more than doubling since 2016 (Statistics Denmark 2024).

From the start, the farmers were organized into five “knowledge exchange groups,” each facilitated by an advisor from a mid-sized, nationally operating agricultural advisory company. For years, the advisory company had been working in interdisciplinary and transdisciplinary approaches with scientists (agronomists and environmental planners) from national universities. In 2022, the Horizon Europe Agroecology-TRANSECT project stimulated the knowledge exchange groups to organize their activities as a reflexive arrangement, with a senior advisor and a connected national university colleague acting as facilitators of the arrangement. To foster connections with the other partners in the project and take the opportunity to reflect on the network’s strategy, the facilitators agreed to apply and evaluate, in their context, the heuristic framework described in this paper.

Data collection

The data collection consisted of three steps: (1) a preparatory exploratory interview with the case-study facilitators and document analysis in month 3 of the Agroecology-TRANSECT project, (2) in-depth semi-structured interviews with key actors in month 5, and (3) online discussion and written feedback with case-study facilitators in month 6.

In step 1, a 30-min exploratory interview was held with the case-study facilitators as part of a project workshop. In a setting around a table, the facilitators were asked to draw up a timeline of what they considered important events for the CA network and to map the current actors and their interrelations. Following the exploratory interview, the case study’s timeline was combined with information from the case study’s action plan for the first project year and a learning history document, both resulting from the co-innovation approach in the project. The resulting updated timeline and actor map were input for the second step and were used to draft a list of potential interviewees. In consultation with one case-study facilitator, two key actors from this list were selected for in-depth interviews next to the two case-study facilitators. Key actors were defined as case-study actors who had a good understanding of the case study’s history and current situation. The key actors should have different roles in the case study to add different perspectives. As part of the preparatory work for step 2, the framework guided a preliminary context description of the case study through document analysis and literature research.

In step 2, additional information on context, actors and their relationships, and barriers and levers was collected using a semi-structured interview guide (Append. 1). The interviewees included the main case-study facilitator, who worked for the agricultural advisory company, a pioneer CA farmer who at the time worked at the same company, a pioneer CA farmer who was part of a knowledge-exchange group, and the second case-study facilitator, who was a university scientist. Three interviews took place in person at the workplaces of the interviewees, and one took place online. Two interviewers were present: one leading the exchange and one with a supportive role. The interviews took 2–3 h each and were recorded.

The in-depth interviews consisted of two parts. First, the updated version of the case study’s timeline was presented to the interviewee. Each interviewee was then asked to describe significant events in the case study’s evolution, either from the timeline or from their own perspective and to elaborate on these. Drawing on the critical incident technique (Gremler 2004), the interviewer asked probing questions aimed at clarifying the interviewees’ perspectives on the barriers, levers, and actors related to the events. In the second part of the interview, the preliminary actor map was discussed with the interviewee. The actors and their connections, discussed in the first part of the in-depth interview, were drawn on the actor map. The interviewee was asked to add or delete actors and connections and to comment on them. The case-study facilitators were additionally asked to clarify aspects of the context that were not clear from the preparatory work.

In step 3, the maps describing (1) actors and their interconnections and (2) barriers and levers, derived from steps 1 and 2 and subjected to a first layer of analysis (i.e., identification of nodes and edges based on coded interview data—see data analysis section), were presented to the case-study facilitators. This step gave them the opportunity to confirm or adjust those maps. Although no major changes were made by the facilitators to the nodes and connections during this step, some clarifications were provided regarding specific terms employed.

Data collection thus resulted in a draft context analysis, and maps describing (1) actors and (2) barriers and levers, with their interconnections.

Data analysis

Interviews from steps 1 and 2 of the data collection were all transcribed and coded with the codes “actors,” “context,” “barriers,” and “levers,” and subcodes were used to group barriers, levers, and their impact (i.e., connections between them) thematically. The information on context was used to enrich the preliminary context description and elaborate the six dimensions (Table 1). Information on actors, barriers, and levers was used in the network analyses described below.

Actor network analysis

The actor network was analyzed using social network analysis (SNA), mixing both quantitative and qualitative methods (Bellotti 2014, Cornu et al. 2023). A growing body of research (e.g., Heath et al. 2009, Edwards 2010, Hollstein 2014, Bellotti 2014, Ahrens 2018, Yousefi Nooraie et al. 2020) highlights the benefits of integrating quantitative and qualitative approaches in SNA to gain a more comprehensive understanding of social phenomena. Quantitative analysis reveals patterns, network structures, and the positions of key actors, and qualitative methods provide rich contextual insights that enhance the interpretation of the quantitative structural metrics.

A social network consists of nodes representing the actors connected by edges, the relations between actors. Actors are represented at organizational level, and the edges are weighted according to the number of interaction types (Append. 2). These interaction types were identified inductively through the analysis of interview data, where any different form of interaction between two actors was coded and assigned a weight of 1. Following an inductive approach, the boundaries of the case-study network were defined according to the perception of the interviewees. Actors and their relationships were included or excluded depending on how relevant interviewees considered them to be. This aligns with what Heath et al. (2009) describe as the realist approach. Each actor considered relevant by at least one interviewee was included in the network.

Drawing on Castella et al. (2022), the actors were categorized according to their roles in the network. The network was analyzed quantitatively using the network metrics degree, weighted degree, and closeness centrality (Table 2). Key actors are those that have a high degree, high weighted degree, or a high closeness centrality. Analysis and visualization of the social network were performed with Gephi (Bastian et al. 2009).

Barriers and levers analysis

The barriers and the levers and their impact were organized as a cognitive map (Garini et al. 2017) by representing them as nodes in a directed network. Cognitive mapping is used to represent individuals’ perceptions related to a particular issue at a given moment in time (ElSawah et al. 2013, Vanwindekens et al. 2013). The edges in the network were classified as either positive (indicating a beneficial influence from a node on another) or negative (indicating a detrimental influence), following the convention of signed networks (Meng et al. 2022). All outgoing arrows from levers have a positive impact on connected nodes, thus reinforcing other levers or mitigating the effects of barriers. In contrast, all outgoing arrows from barriers exert a negative impact on connected nodes, reinforcing the negative effect of other barriers or reducing the positive influence of levers that are connected to the node. The initial network of barriers and levers was drafted by the principal investigator of the study based on the thematic sub-codes derived from the exploratory (step 1) and in-depth interviews (step 2) described in data collection section. This network was then confirmed and refined with the case-study facilitators (step 3), before being analyzed quantitatively and visualized with the Gephi software (Bastian et al. 2009).

Similar to actor network analysis, network metrics can be used to investigate the structure of the network and gain insight into the role and importance of individual barriers and levers (Vanwindekens et al. 2014). The metrics’ out-degree, (positive or negative) in-degree, in-degree balance and betweenness centrality were calculated to analyze the network of barriers and levers (Table 2).

A key barrier has a low positive in-degree, a neutral or negative in-degree balance, a high out-degree or a high betweenness. A key lever has low negative in-degree, a neutral or positive in-degree balance, a high out-degree or a high betweenness. Using the metrics, the key barriers and levers were classified into initiative-specific types, reflecting their position in the network. The key barriers were classified as blocking, recurring, and eased barriers (Table 3). The key levers were classified as powerful, influential, connecting, and minor levers (Table 3).

RESULTS

Context of the Danish case study

The context framework (Table 1) highlighted a highly technologized, export-oriented agricultural production sector under strict environmental policies. The biophysical landscape is dominated by agricultural land use on which mostly fodder crops are grown. The main agricultural products are pork and dairy. Most of the farms exceed 100 ha and operate with high debts and low margins. Large food companies that originated from farmer cooperatives are dominating the agricultural market. Denmark is among the countries with the highest consumption of organic products.

Case-study actors and their network

In total, 27 actors were identified, including the three actors constituting the core of the reflexive arrangement (advisory company (A1; Fig. 1), CA farmers (A2) and researchers from National University A (A3)). Among the 27 actors, one played an advisory role (A1), another was directly involved in farming (A2), 10 were engaged in influencing or developing agricultural policies (A4, A5, A9, A10, A12, A13, A14, A15, A16, A17), eight had predominant economic impacts and interests (A8, A18, A20, A21, A22, A23, A24, A25), six were linked to research (A3, A6, A7, A11, A26, A27), and one was categorized under society (A19). The 27 actors, their characterization in terms of the metrics degree, weighted degree, and closeness centrality, as well as the relationships among them, are illustrated in Fig. 1.

The agricultural advisory company (A1) appears as the most important actor, with the highest values for degree (19), weighted degree (37), and closeness centrality (0.79). Conservation agriculture farmers (A2) are highly connected (degree of 7) but more distant to other actors than the agricultural advisory company (A1). National University A (A3) has fewer (degree of 5) but strong relations (weighted degree of 14), indicated by the high weighted degree in comparison to the degree. Other key actors are related to policy or research. Key policy actors are more strongly related and closer to non-research actors. They include the Danish Agriculture and Food Council (A12), the European Union (A10), the Danish parliament (A5), the national farmer association for reduced tillage (A9), and Danish ministries (A4). Key research actors are National University B (A7), the non-profit research center (A11), and National University C (A6).

Barriers, levers, and their interrelations

We found, in total, 30 barriers and 35 levers (Fig. 2) (see Append. 3 for a comprehensive list and description of the barriers and levers), from which 11 key barriers and 10 key levers were identified. The key barriers included 5 blocking, 2 recurring, and 4 eased barriers, whereas the key levers included 1 powerful, 3 influential, 2 connecting, and 4 minor levers (Table 4).

The lack of practical knowledge about CA in Denmark (B9) (Fig. 2A) is the barrier with the highest betweenness, indicating a connecting role in the network. Its reinforcing effect on the risk of yield reduction due to CA (B8) and on the difficulty for advisors and scientists to leave the expert role (B4) are outweighed by six levers: the access to knowledge through social media (L25), knowledge-exchange groups enabling farmers to be the source of CA development (L34), knowledge-sharing between farmers (L26), knowledge-exchange groups enabling to build trust and share experiences honestly (L8), field demonstrations (L28), and the contact with agricultural experts being reassuring when trying something new (L24). The high in-degree balance of 6 indicates that the barrier of lack of practical knowledge is potentially overcome, which is in line with a statement from one of the interviewed farmers “I believe the farmers can fix that [practical problems with CA]. We can fix that in the knowledge groups.”

Knowledge sharing between farmers (L26) connects different ways of knowledge production (Fig. 2A). Knowledge-exchange groups enhance knowledge sharing by enabling their members to build up trust and share experiences honestly (L8) and to be the source of CA development (L34), even though conflicts due to different mindsets (B2) can inhibit the members from building mutual trust. The lack of formal knowledge collection and reporting (B23) limits knowledge sharing (L26), whereas it is enhanced by the newsletter of the agricultural advisory company (L27). Furthermore, the Healthy Soil conference enhances knowledge sharing by providing farmers a platform to discuss CA (L4) and motivates farmers with stories of inspiring CA farmers who show that CA works (L23). Even though the Healthy Soil Conference has been successful in attracting a large number of farmers, it is challenging to keep it interesting for frontrunners (B7), who are key for knowledge sharing among farmers (L26). Also, field demonstrations (L28) address the lack of practical knowledge (B9) and make the agricultural advisory company more attractive for CA farmers (L11). The field demonstrations showed the technical aspects and machinery, influenced by the focus of farmers on yield and big machinery (B24).

National University A’s interdisciplinary and transdisciplinary approach (L31) (Fig. 2B) is an influential lever, facilitating the collaboration of social scientists with advisors (L1) and farmers (L3). Thereby, it enabled the advisory company (A1) to join a European research and innovation action project, also addressing the difficulty of getting funding (B26). The interviews indicated that the approach of National University A (L31) is valued by the EU (L5) but, besides a grant from a foundation (L19), not much supported at the national level where traditional natural science approaches dominate in agricultural transition research (B2). The collaboration of social scientists and advisors (L1) enhances the collaboration of social scientists and farmers (L3), with advisors as intermediaries. This opportunity comes with constraints, as advisors tend to self-censor by keeping information on novel approaches that they consider too risky and potentially harmful for the (economic) relationship away from their customers. Thereby the advisory company is limited in innovation through novel approaches that do not fulfil their customers’ expectations (B5), a blocking barrier. Also impeding the emerging collaborations of social scientists with advisors and farmers are traditions that shape the aims and expectations of the partners (B3), the overcoming of which requires building trust (L32) through personal interactions over time. The collaboration of social scientists and advisors (L1) emerges as a minor lever with the potential to advance a facilitative advisory role (L2) and to overcome the barrier of the advisors’ lacking training in facilitation skills and co-creation with farmers (B6). The difficulty of leaving the expert role for advisors and scientists (B4) constitutes a strong blocking barrier that also impacts the collaboration of social scientists and farmers (L3) and advisory companies’ limited innovativeness (B5). Moving away from the expert role will be necessary, considering the context specificity of CA and the limited practical knowledge about CA in Denmark (B9). Still, farmers’ expectations of receiving advice and services rather than being included in strategic and operational knowledge development make it difficult.

Farmers’ focus on yield and big machinery (B19) (Fig. 2C), is a blocking barrier that is reinforced by masculinity in agriculture (B21) and farmers’ education, in which ploughing is taught as a central part of farming (B28). The fascination with soil and soil life that many CA farmers share (L7) has potential to shift farmers’ focus and contributes to stories of inspiring CA farmers who show that CA works (L25). The predominant farmers’ focus on yield and big machinery (B19) reinforces the recurring barrier of risking a yield reduction due to CA (B8). This risk is especially high during the conversion to CA, when context-specific knowledge on the application of CA still has to be acquired, and the beneficial effects of CA are not yet occurring, such as the increase in the farm’s environmental robustness (L22) through reduced erosion and higher drought resilience stabilizing yields in the long-term. Even though farming costs are reduced (L9), according to the interviews even to the point that yield losses are compensated, the financial pressure on farms (B29) makes farmers reluctant to take risks. A focus on economic benefits (B29) potentially limits the success of stories of inspiring CA farmers (L23).

The blocking barrier Contested knowledge (B1) (Fig. 2D) is reinforced by the blocking barrier B2: The dominance of National University B’s natural science approach in policy making. Both barriers impact the barrier of the current legislation not supporting CA (B11). This recurring barrier is also enforced by the agricultural council representing the interests of major companies and the majority of farmers (B14) and thereby not being interested in supporting CA. Approaches to bring CA farmers in contact with politicians (L16) and the farmer association for reduced tillage advocating CA farmers’ interests politically (L15) constitute levers to bring about legislative support for CA. Carbon dioxide certificates for CA (L17) represent a minor lever (L17) with the potential to generate direct additional income (B16) from CA but are blocked by contested knowledge (B1) and CA not being considered in the value of the land (B15).

Conservation agriculture is lacking visibility in society (B10) (Fig. 2E), which reinforces other barriers: the current legislation that does not support CA (B11), the use of CA not being considered in the value of land (B15), and the lack of remuneration for adopting CA (B16). The lack of visibility of CA in society (B10) is eased by the collaboration with nature NGOs (L13) and the Thinktank (L14), and the promotion of CA by connecting it to food (L12). The collaboration with the nature NGOs (L13) evolved from the advisory company inviting nature NGOs to give a speech for farmers at the company’s annual Healthy Soil conference (L33) and has the potential to promote legislative support for CA (B11). Besides enhancing visibility, the collaboration with nature NGOs (L13) and the Thinktank (L14) can also further the understanding of CA in society (B13). The collaboration with the Thinktank (L14) additionally provides an opportunity to get into contact with young farmers (L20).

DISCUSSION

We developed a heuristic framework to portray agroecological initiatives at the start of reflexive arrangements. The framework captures the context, the actor network, and the barriers and levers for the development of the agroecological initiative. The novelty of the framework resides in the combination of elements and in its heuristic rather than comprehensive purpose to fit its use in a transdisciplinary research context. The application of the framework to the case of the Danish CA network highlighted a highly technologized, export-oriented agricultural sector, with a focus on feed crops in arable farming, producing under strict environmental policies, with many farms operating with high debts and low margins. Next to the three actors that are part of the reflexive arrangement, key actors were related to policy and research. In comparison with researchers, policy actors were more strongly related to and embedded in the network. Connections with economy actors existed, but they were loosely related to the network, while society actors were almost non-present. Key barriers and levers comprised a broad range of themes, such as the role of advisory actors and scientists, the mobilization of horizontal knowledge structures and the lack of financial reward and visibility for CA.

In this section, we discuss the framework in relation to its aim of providing a starting point for scientist–societal actor collaborations in reflexive arrangements. The results for the Danish CA network are elaborated using the multi-level perspective. The state of technological, network, and institutional anchoring is discussed using the main themes that emerged from the context-related analysis of actors and barriers and levers. We describe how the results may influence the development of the reflexive arrangement.

Portraying agroecological initiatives as a starting point in reflexive arrangements

This study was inspired by the need to establish working relations between societal actors in agroecological initiatives and scientists, collaborating for 4 yr in a European research and innovation project. To establish connections between scientific capabilities and the development status of the agroecological initiatives, a methodology was needed that balanced scientific rigor, salience for the users, and timeliness. Scientific rigor of the framework resides in the constituent elements that were selected from various methods proposed to characterize or map socio-ecological systems. Although the context characterization sketches the setting of agroecological initiative in the overarching agri-food system, the networks of actors, barriers, and levers provide actionable knowledge (Geertsema et al. 2016) by focusing on the agroecological initiative and its history. Salience for the users was achieved by data collection in various rounds, providing feedback on results and asking for user input on credibility of results and relevance for the agroecological initiative. These cycles of data collection, analysis, and reflection were part of the overarching project’s learning-oriented approach that built on earlier co-innovation approaches of the team (Rossing et al. 2021, 2023). Thus, developing the agroecological initiative’s portrait was one of the means of building social capital in the project. We estimate that data collection and analysis took around 2 mo, allowing relatively fast scientific input into the innovation dynamics of the agroecological initiative.

Insights for the Danish conservation agriculture network

From a multi-level perspective, agroecological initiatives may be seen as niches that are external to the regime, characterized by a divergent structure and alternative values compared with the industrialized agriculture regime (Levidow et al. 2014, Morel et al. 2020). The Danish case study, in contrast, shared values with the dominant regime, such as the focus on high yields and the use of pesticides, whereas other values, such as the care for beneficial insects by avoiding use of insecticides (Hansen et al. 2020) and for soil quality by reducing or abstaining from tillage differed radically. The case study thus constituted a niche in the regime, which Elzen et al. (2012) refer to as a hybrid.

An important goal of the Danish case study was to make CA mainstream, i.e., to anchor CA in the regime to achieve, for example, sector targets for climatic mitigation (Ministry of Food, Agriculture and Fisheries 2021) and nitrogen leaching reduction (Environment Agency 2023). Three types of anchoring have been distinguished. Technological anchoring occurs when technical characteristics of an innovation become defined by involved actors. Network anchoring refers to an expansion or intensification of the network of actors that support CA practices. Institutional anchoring means the development of new rules related to CA practices, which can be cognitive, normative, or economic (Elzen et al. 2012).

Technological anchoring: advisors as knowledge facilitators

Technological anchoring appeared in the development and sharing of practical CA knowledge. The number and diversity of levers addressing lack of knowledge as a key barrier was found to be large, and one of the interviewees concluded that the technical difficulties could be overcome on-farm. The key levers, all initiated by the advisory company (A1), facilitated knowledge sharing between farmers and included the organization of the Healthy Soil Conference as a platform for farmers to discuss CA, demonstrations, and knowledge-exchange groups providing safe spaces for farmers to learn and experiment. This focus on horizontal knowledge structures through a facilitative and participatory advisory approach was distinct from the dominant centralized knowledge production and top-down knowledge diffusion (Anderson et al. 2019a).

The traditional, top-down advisory role has been questioned as to whether it effectively addresses current challenges in agriculture (Landini et al. 2021, Krafft et al. 2022), as it neglects the complexity of systems and their context specificity (Charatsari et al. 2019). This is supported by earlier studies involving some of the farmers in the network (Hansen et al. 2020). To support agroecological practices through the use of CA principles, more systemic, facilitative, and participatory approaches of advisory actors are needed (Heleba et al. 2016, Charatsari et al. 2019, Landini et al. 2021, Krafft et al. 2022). Such an advisory approach strengthens horizontal knowledge structures (Anderson et al. 2019a, Bourne et al. 2021) and enhances the development of farmer skills to solve complex problems arising in their specific context (Cristofari et al. 2017, Charatsari et al. 2019, Bourne et al. 2021, Krafft et al. 2022), which makes it more effective (Ataei et al. 2019, Anderson et al. 2019a).

The facilitative and participatory approaches in the case study were hampered by the dominant regime. Our analysis identified as key barriers the traditional role of advisors and the risk of customer loss associated with the new ways of operating. To overcome such limitations, Krafft et al. (2022) point out the importance of advisors’ skills, interdisciplinary collaboration, and farmer engagement in discussions. The development of advisors’ facilitation skills in their evolving role requires time and support from diverse disciplines, notably facilitated through collaboration with National University A (A3).

In summary, the advisory company at the core of the case study facilitated knowledge development and sharing among farmers through a variety of activities, including workshops, field demonstrations, and conferences, thereby enhancing farmers’ skills to solve complex problems. This appeared to have overcome a lack of knowledge as a barrier, indicative of the successful technological anchoring process. Although successful, the advisory actor’s approach to strengthening horizontal knowledge structures remained a niche within the overarching dominating top-down advisory structure.

Network anchoring: advisors as network constructors

Network anchoring is evident in the collaboration of the Danish case study with regime-related actors, including hybrid actors who are part of the regime but hold differing views (Elzen et al. 2012, Diaz et al. 2013).

The network analysis and the analysis of barriers and levers revealed the dominance of regime actors related to policy and research and their blocking effect for anchoring. Key policy actors shaped the dominant regime of public policies and political power, by which CA is not supported. Key research actors reinforced the dominant regime of centralized knowledge production and top-down knowledge diffusion, which hampered the mobilization of horizontal knowledge structures in the case study.

As the most central and connected actor in the network, the advisory company held a crucial position for network anchoring to progress the agroecological transition (Heleba et al. 2016, Bourne et al. 2021, Krafft et al. 2022). The advisory company recognized the potential of relations with hybrid actors in connecting with the Thinktank and nature NGOs, as indicated by key levers. The Thinktank included regime actors in dialogs but aimed to disrupt dominant discourses. The nature NGOs were embedded in the regime structures but questioned regime values. The ties between the agricultural advisory and nature NGOs were especially novel, considering their commonly different perspectives. Hybrid actors related to society and economy were either loosely related or non-present in the network and may provide useful entry points for enhancing network anchoring.

In summary, the advisory company took the role of network constructor. To overcome the dominance of regime actors, enhancing the building of connections with hybrid actors is promising for network anchoring.

Institutional anchoring: contested knowledge as a barrier

Conservation agriculture in Denmark largely lacked institutional support. The analysis brought out contested knowledge as a key barrier to agreement on the relevance of CA for C sequestration. Beneficial, less contested aspects of CA (Farooq and Siddique 2015) were only highlighted to a very limited degree. Researchers from National University B (A7) concluded that data from a national long-term experiment did not show significant differences in soil organic C concentrations in the 0–50 cm soil profile when comparing direct drilling and ploughing and did not fully confirm a positive effect of straw retention on soil organic C content (Gómez-Muñoz et al. 2021). The experiment, established in 2002 at two different research farm sites, combined four tillage treatments with four crop rotations and ways of straw management in a split-plot design (Hansen et al. 2010, 2015). The findings did not match with practitioners’ perceived changes of their soil after transition to CA. Proponents of CA questioned whether research farm plots were suitable to draw conclusions about the effects of farmers’ situated CA practices. Successful CA application lies in the combination of adapting the three CA principles to the local context, which is not the case in standardized treatments in plot experiments (Rodenburg et al. 2020). Plot experiments, such as the Danish long-term research trial, produce generalizable agronomic insight but fail to capture the situated complexity of activities and interactions related to farming practices (Lacoste et al. 2022). Additionally, they are difficult for practitioners to relate to (Hansen et al. 2020), indicating that the effects of CA observed on CA farms need to be measured in that context. The relevance of on-farm experiments was recently emphasized by Lacoste et al. (2022), pointing out how the engagement with farming realities creates value for both scientists and farmers. Engaging in on-farm experiments, however, challenges current scientific approaches such as dealing with variability of farmer management, requiring more frequent communication between researchers and farmers, and maybe also different cross-disciplinary analytical tools (González-Sánchez et al. 2012, Anderson et al. 2019a).

Situations like the disagreement about the effect of CA on C sequestration have been described as part of wicked problems. Wicked problems are characterized by a lack of agreement on problem definition, i.e., the contextualized nature of CA effects on C sequestration, due to conflicting values and interests on the one hand, and by uncertainty of knowledge on proposed solutions on the other, i.e., the effect of size of the experimental treatments (Xiang 2013). Uncertainty about ecological processes and conflicting social values have been found to be a breeding ground of wicked problems in socio-ecological systems (Norris et al. 2016). The results of the on-station experiments were reported in a white paper (Munkhom et al. 2020) that was accepted by government as a basis for policy making, thus reinforcing the difficulty of institutional anchoring. A scientifically validated objectification of the performance of contextualized on-farm CA practices, compared with on-station experiments, may provide an avenue to overcome the current stalemate. Nevertheless, from a scientific perspective, a shift in the focus is needed, by not only looking at a single factor of the practice, such as C sequestration, but rather restructuring farmer–researcher relationships and addressing complexity and uncertainty through joint farm system exploration.

Methodological considerations

The actor network required simplification by representing actors in terms of organizations, thereby overlooking relationships between individuals. Including the degree of human agency that individual actors exhibit could significantly strengthen the relevance of the results (Gaitán-Cremaschi et al. 2022). We considered the simplification of the actor network acceptable as it enhanced the clarity of representation, and crucial human agency of individual actors was captured in the barriers and levers (e.g., the pioneer CA farmer working in advisory capacity (L30)). Nevertheless, considering the level of individuals in organizations is likely a necessary next step for use of the results.

The weights of the edges of the actor network were determined as the number of interaction types mentioned in the interviews, with no consideration of the intensity of each type of interaction. Also, the network of barriers and levers does not currently account for the intensity of the relationships between nodes or the initial magnitude of each barrier and lever individually. This limitation may be overcome by introducing an additional step, in which key actors themselves assign weights to nodes and relationships.

Additionally, it is important to emphasize that this “portrait” is guided by the perceptions of four key case-study actors and shaped by the analysts’ interpretation. Therefore, the potential for incompleteness as well as interviewees’ and researchers’ bias constitute potential threats to the validity of the findings. Assessing validity depends on the types of knowledge claims and methods used, involves trade-offs between different threats, and requires judgment based on background knowledge (Hammersley 2008). In this study, the focus on reaching a shared understanding of the case study as a starting point for a reflexive arrangement has guided the development of our methods. Rather than aiming for time-consuming comprehensive network analysis, the result of our framework provides contextualized hypotheses for elaboration during the reflexive arrangement. The initiative’s portrait is considered a tool for further investigation that can be continuously updated and reinterpreted based on the feedback and perspectives of the participants. The lack of comprehensiveness and possible biases are thus purposefully traded off with the timeliness and actionability of the knowledge generated. At the same time, efforts were made to reduce the threats on the validity of the findings. First, reliability of the conceptual model derived from a combination of literature sources, resulted in theoretical triangulation based on three pillars (i.e., context, actor network, and barriers and levers); reliability of the data base was enhanced by data source triangulation—through interviews with different stakeholders. Second, the different rounds of data collection and analysis helped ensure data accuracy through member-checking (Lincoln and Guba 1985, as cited in Mabry 2008), where key participants from the case study were asked during interviews to confirm, elaborate on, and, where needed, refute the data and interpretations. Given that validity is always a matter of judgment based on background knowledge (Hammersley 2008), we assume that case-study participants are well-positioned to assess the validity of research findings. Finally, rather than aiming to produce grand grounded theory, we instead seek local theory (Mabry 2008) at the scale of the specific case study (i.e., the Danish CA Network), in line with the aim of alignment in reflexive arrangements.

Finally, it would be valuable to explore how alternative approaches to studying complex phenomena might offer new perspectives, using different system concepts. This could complement—or reshape—the portrait of the Danish CA Network derived according to the three pillars of our framework. For instance, whereas the actor network analysis primarily focuses on human actors and their interactions, innovation scholars have also recognized the critical role non-human actors play in innovation networks (Jarrahi and Sawyer 2019, Granstrand and Holgersson 2020). In this context, investigating how applying actor-network theory to conceptualize socio-technical assemblages (Jarrahi and Sawyer 2019) could present a promising analytical alternative.

CONCLUSION

Drawing on more elaborate characterization and assessment approaches, we proposed a learning-oriented framework to develop a shared portrait of an agroecological initiative as a starting point for a reflexive arrangement. This framework is illustrated through its application to a real-world farm management initiative, namely the Danish CA network. Relying on interviews with a limited but diverse set of actors from the initiative, comprehensiveness is deliberately sacrificed for the purpose of timeliness and actionability of the results, without compromising scientific rigor. The Danish CA network facilitators commented how the analytical results helped them to see the position of their initiative differently. In particular, they mentioned the perspective of CA as a niche in the regime and the lack of connections to actor groups that could enhance the visibility and recognition of CA farmers’ positive contributions. They also emphasized the innovative development of horizontal networks by the advisory company and the need to assess objectively CA performance on-farms. Such reframing of “how we see the world” has been denoted as social learning, which is considered by many scholars essential for transformative change.

The process of applying the framework, i.e., having face-to-face working sessions at group and individual levels, online feedback, and frequent questioning by the researchers of the facilitators, contributed to trusting working relationships among the participants. The credibility and transparency of the approach, the attention to local details, the respectful use of the information provided by the actors, and the salience of the results all contributed to social capital, which is another important element in transformative change.

This study is an example of how research can be designed inclusively for the purpose of answering how-to questions associated with transformative change. Comparative analysis of the several agroecological initiatives that are part of the overarching Agroecology-TRANSECT project using the approach presented is expected to result in actionable knowledge for the initiatives, as well as enhance learning across transformative efforts.

LITERATURE CITED

Adhikari, K., A. E. Hartemink, B. Minasny, R. Bou Kheir, M. B. Greve, and M. H. Greve. 2014. Digital mapping of soil organic carbon contents and stocks in Denmark. PLoS ONE 9(8):e105519. https://doi.org/10.1371/journal.pone.0105519

Agroecology Europe. 2020. Agroecology initiatives in European countries. Agroecology Europe Youth Network, Brussels, Belgium. https://www.agroecology-europe.org/wp-content/uploads/2020/11/AEEU-Mapping-Report-agroecological-practices-November-version.pdf

Ahrens, P. 2018. Qualitative network analysis: a useful tool for investigating policy networks in transnational settings? Methodological Innovations 11(1):205979911876981. https://doi.org/10.1177/2059799118769816

Alvarez, S., C. J. Timler, M. Michalscheck, W. Paas, K. Descheemaeker, P. Tittonell, J. A. Andersson, and J. C. J. Groot. 2018. Capturing farm diversity with hypothesis-based typologies: an innovative methodological framework for farming system typology development. PLoS ONE 13(5):e0194757. https://doi.org/10.1371/journal.pone.0194757

Anderson, C. R., J. Bruil, M. J. Chappell, C. Kiss, and M. P. Pimbert. 2019a. From transition to domains of transformation: getting to sustainable and just food systems through agroecology. Sustainability 11(19):5272. https://doi.org/10.3390/su11195272

Anderson, C. R., C. Maughan, and M. P. Pimbert. 2019b. Transformative agroecology learning in Europe: building consciousness, skills and collective capacity for food sovereignty. Agriculture and Human Values 36(3):531-547. https://doi.org/10.1007/978-3-031-19400-9_2

Arler, F., M. S. Jørgensen, D. Galland, and E. M. Sørensen. 2015. Prioritering af fremtidens arealanvendelse i Danmark. Institute for Planning, Aalborg University, Copenhagen, Denmark. https://tekno.dk/app/uploads/2015/08/Prioritering-af-fremtidens-arealanvendelse-i-Danmark.pdf

Ataei, P., H. Sadighi, M. Chizari, and E. Abbasi. 2019. Analysis of farmers’ social interactions to apply principles of conservation agriculture in Iran: application of social network analysis. Journal of Agricultural Science and Technology 21(Supp.):1657-1671. https://jast.modares.ac.ir/article-23-21426-en.pdf

Averbuch, B., M. Hvarregaard Thorsøe, and C. Kjeldsen. 2021. Longue durée study of agricultural transitions in Denmark using multi-level perspective. Geografisk Tidsskrift-Danish Journal of Geography 121(1):30-45. https://doi.org/10.1080/00167223.2021.1886958

Averbuch, B., M. H. Thorsøe, and C. Kjeldsen. 2022. Using fuzzy cognitive mapping and social capital to explain differences in sustainability perceptions between farmers in the northeast US and Denmark. Agriculture and Human Values 39(1):435-453. https://doi.org/10.1007/s10460-021-10264-4

Barrios, E., B. Gemmill-Herren, A. Bicksler, E. Siliprandi, R. Brathwaite, S. Moller, C. Batello, and P. Tittonell. 2020. The 10 elements of agroecology: enabling transitions towards sustainable agriculture and food systems through visual narratives. Ecosystems and People 16(1):230-247. https://doi.org/10.1080/26395916.2020.1808705

Barzman, M., and S. Dachbrodt-Saaydeh. 2011. Comparative analysis of pesticide action plans in five European countries. Pest Management Science 67(12):1481-1485. https://doi.org/10.1002/ps.2283

Bastian, M., S. Heymann, and M. Jacomy. 2009. Gephi: an open source software for exploring and manipulating networks. Proceedings of the International AAAI Conference on Web and Social Media 3(1):361-362. https://doi.org/10.1609/icwsm.v3i1.13937

Bellotti, E. 2014. Qualitative networks: mixed methods in sociological research. Routledge, London, UK.

Blanch-Ramirez, J., L. Calvet-Mir, L. Aceituno-Mata, and P. Benyei. 2022. Climate change in the Catalan Pyrenees intersects with socioeconomic factors to shape crop diversity and management. Agronomy for Sustainable Development 42(5):91. https://doi.org/10.1007/s13593-022-00806-3

Bos, A. P., P. W. G. Groot Koerkamp, J. M. J. Gosselink, and S. Bokma. 2009. Reflexive interactive design and its application in a project on sustainable dairy husbandry systems. Outlook on Agriculture 38(2):137-145. https://doi.org/10.5367/000000009788632386

Bourne, M., L. L. de Bruyn, and J. Prior. 2021. Participatory versus traditional agricultural advisory models for training farmers in conservation agriculture: a comparative analysis from Kenya. The Journal of Agricultural Education and Extension 27(2):153-174. https://doi.org/10.1080/1389224X.2020.1828113

Britt, H., and R. Wilson-Grau. 2012. Outcome harvesting. Ford Foundation, Middle East and North Africa Office, Cairo, Egypt. https://outcomeharvesting.net/wp-content/uploads/2016/07/Outcome-Harvesting-Brief-revised-Nov-2013.pdf

Castella, J.-C., G. Lestrelin, S. Phimmasone, H. Tran Quoc, and P. Lienhard. 2022. The role of actor networks in enabling agroecological innovation: lessons from Laos. Sustainability 14(6):3550. https://doi.org/10.3390/su14063550

Charatsari, C., E. D. Lioutas, A. Papadaki-Klavdianou, A. Koutsouris, and A. Michailidis. 2019. Experiential, social, connectivist, or transformative learning? Farm advisors and the construction of agroecological knowledge. Sustainability 14(4):2426. https://doi.org/10.3390/su14042426

Climate Data. n.d. Climate: region Zealand. AM Online Projects, Oedheim, Germany. https://en.climate-data.org/europe/denmark/region-zealand-425/

Cornu, M.-A., R. Frick, I. R. Chongtham, I. Iocola, S. Canali, L. Colombo, P. Radzikowski, J. Stalenga, L. Viguier, D. Drexler, A. Schneider, D. Stilmant, and F. M. Vanwindekens. 2023. Identification and description of relationships between actors involved in crop diversification experiences across Europe. Agronomy for Sustainable Development 43(5):67. https://doi.org/10.1007/s13593-023-00906-8

Coupaye, L. 2015. Chaînes opératoire, transect et théories. Quelques réflexions sur le parcours d’une méthode classique. Pages 69-84 in A. Leroi-Gourhan. L'homme, tout simplement. Boccard, Paris, France.

Cristofari, H., N. Girard, and D. Magda. 2017. Supporting transition toward conservation agriculture: a framework to analyze the learning processes of farmers. Hungarian Geographical Bulletin 66(1):65-76. https://doi.org/10.15201/hungeobull.66.1.7

Cronin, E., A. Fieldsend, E. Rogge, and T. Block. 2022. Multi-actor Horizon 2020 projects in agriculture, forestry and related sectors: a multi-level innovation system framework (MINOS) for identifying multi-level system failures. Agricultural Systems 196:103349. https://doi.org/10.1016/j.agsy.2021.103349

Denver, S., J. Nordström, and T. Christensen. 2019. Is an increase in organic consumption accompanied by a healthier diet? A comparison of changes in eating habits among Danish consumers. Journal of Food Products Marketing 25(5):479-499. https://doi.org/10.1080/10454446.2019.1600449

Diaz, M., I. Darnhofer, C. Darrot, and J.-E. Beuret. 2013. Green tides in Brittany: what can we learn about niche-regime interactions? Environmental Innovation and Societal Transitions 8:62-75. https://doi.org/10.1016/j.eist.2013.04.002

Douthwaite, B., and E. Hoffecker. 2017. Towards a complexity-aware theory of change for participatory research programs working within agricultural innovation systems. Agricultural Systems 155:88-102. https://doi.org/10.1016/j.agsy.2017.04.002

Edwards, G. 2010. Mixed methods approaches to social networks analysis. ESRC National Centre for Research Methods, Southampton, UK.

Egmose, J., S. G. Jacobsen, H. Hauggaard-Nielsen, and L. Hulgård. 2021. The regenerative turn: on the re-emergence of reciprocity embedded in living ecologies. Globalizations 18(7):1271-1276. https://doi.org/10.1080/14747731.2021.1911508

ElSawah, S., A. Mclucas, and J. Mazanov. 2013. Using a cognitive mapping approach to frame the perceptions of water users about managing water resources: a case study in the Australian Capital Territory. Water Resources Management 27(9):3441-3456. https://doi.org/10.1007/s11269-013-0357-5

Elzen, B., and B. Bos. 2019. The RIO approach: design and anchoring of sustainable animal husbandry systems. Technological Forecasting and Social Change 145:141-152. https://doi.org/10.1016/j.techfore.2016.05.023

Elzen, B., B. van Mierlo, and C. Leeuwis. 2012. Anchoring of innovations: assessing Dutch efforts to harvest energy from glasshouses. Environmental Innovation and Societal Transitions 5:1-18. https://doi.org/10.1016/j.eist.2012.10.006

Engel, P. G. H. 1997. The social organization of innovation: a focus on stakeholder interaction. Royal Tropical Institute, Amsterdam, The Netherlands.

Environment Agency. 2023. Vandområdeplanerne 2021-2027. Danish Agency for Green Land Redevelopment and the Aquatic Environment, Copenhagen, Denmark. https://mst.dk/erhverv/rent-miljoe-og-sikker-forsyning/vandmiljoe/vandomraadeplaner/overblik-vandomraadeplanerne-2021-2027/vandomraadeplanerne-2021-2027

Escobar, N., N. J. Romero, and C. I. Jaramillo. 2019. Typology of small producers in transition to agroecological production. Agronomy 17(6):2242-2259. https://doi.org/10.15159/AR.19.221

European Commission. 2020. Farm to fork strategy for a fair, healthy and environmentally-friendly food system. European Commission, Brussels, Belgium.

Farooq, M., and K. H. M. Siddique. 2015. Conservation agriculture. Springer International Publishing, Cham, Switzerland. https://doi.org/10.1007/978-3-319-11620-4

Fieldsend, A. F., E. Cronin, E. Varga, S. Biró, and E. Rogge. 2021. “Sharing the space” in the agricultural knowledge and innovation system: multi-actor innovation partnerships with farmers and foresters in Europe. The Journal of Agricultural Education and Extension 27(4):423-442. https://doi.org/10.1080/1389224X.2021.1873156

Fischer, J., and M. Riechers. 2019. A leverage points perspective on sustainability. People and Nature 1(1):115-120. https://doi.org/10.1002/pan3.13

Food and Agriculture Organization (FAO). 2018. The 10 elements of agroecology: guiding the transition to sustainable food and agricultural systems. FAO, Rome, Italy.

Food and Agriculture Organization (FAO). 2022. Conservation Agriculture. FAO, Rome, Italy.

Gaitán-Cremaschi, D., L. Klerkx, N. Aguilar-Gallegos, J. Duncan, A. Pizzolón, S. Dogliotti, and W. A. H. Rossing. 2022. Public food procurement from family farming: a food system and social network perspective. Food Policy 111:102325. https://doi.org/10.1016/j.foodpol.2022.102325

Garini, C. S., F. Vanwindekens, J. M. S. Scholberg, A. Wezel, and J. C. J. Groot. 2017. Drivers of adoption of agroecological practices for winegrowers and influence from policies in the province of Trento, Italy. Land Use Policy 68:200-211. https://doi.org/10.1016/j.landusepol.2017.07.048

Geels, F. W., and J. Schot. 2007. Typology of sociotechnical transition pathways. Research Policy 36(3):399-417. https://doi.org/10.1016/j.respol.2007.01.003

Geertsema, W., W. A. Rossing, D. A. Landis, F. J. Bianchi, P. C. van Rijn, J. H. Schaminée, T. Tscharntke, and W. van der Werf. 2016. Actionable knowledge for ecological intensification of agriculture. Frontiers in Ecology and the Environment 14(4):209-216. https://doi.org/10.1002/fee.1258

Ghosh, B., V. Shaw, E. Zinkstock, and E. Steinmueller. 2022. Pentagonal map for system analysis (individual version). TIP Resource Lab, Transformative Innovation Policy Consortium, University of Sussex Business School, Brighton, UK. https://tipresourcelab.net/resource/pentagonal-map-for-system-analysis-individual-version/

Giller, K. E., R. Hijbeek, J. A. Andersson, and J. Sumberg. 2021. Regenerative agriculture: an agronomic perspective. Outlook on Agriculture 50(1):13-25. https://doi.org/10.1177/0030727021998063

Gliessman, S. 2016. Transforming food systems with agroecology. Agroecology and Sustainable Food Systems 40(3):187-189. https://doi.org/10.1080/21683565.2015.1130765

Gómez-Muñoz, B., L. S. Jensen, L. Munkholm, J. E. Olesen, E. Møller Hansen, and S. Bruun. 2021. Long-term effect of tillage and straw retention in conservation agriculture systems on soil carbon storage. Soil Science Society of America Journal 85(5):1465-1478. https://doi.org/10.1002/saj2.20312

González-Sánchez, E. J., R. Ordóñez-Fernández, R. Carbonell-Bojollo, O. Veroz-González, and J. A. Gil-Ribes. 2012. Meta-analysis on atmospheric carbon capture in Spain through the use of conservation agriculture. Soil and Tillage Research 122:52-60. https://doi.org/10.1016/j.still.2012.03.001

Granstrand, O., and M. Holgersson. 2020. Innovation ecosystems: a conceptual review and a new definition. Technovation 90-91:102098. https://doi.org/10.1016/j.technovation.2019.102098

Gremler, D. D. 2004. The critical incident technique in service research. Journal of Service Research 7(1):65-89. https://doi.org/10.1177/1094670504266138

Grivins, M., M. H. Thorsøe, and D. Maye. 2021. Financial subjectivities in the agricultural sector: a comparative analysis of relations between farmers and banks in Latvia, Denmark and the UK. Journal of Rural Studies 86:117-126. https://doi.org/10.1016/j.jrurstud.2021.06.006

Guzmán, G. I., D. López, L. Román, and A. M. Alonso. 2013. Participatory action research in agroecology: building local organic food networks in Spain. Agroecology and Sustainable Food Systems 37(1):127-146. https://doi.org/10.1080/10440046.2012.718997

Hammersley, M. 2008. Assessing validity in social research. Pages 42-53 in P. Alasuutari, L. Bickman, and J. Brannen, editors. The SAGE handbook of social research methods. SAGE Publications, London, UK. https://doi.org/10.4135/9781446212165.n4

Hansen, E. M., L. J. Munkholm, B. Melander, and J. E. Olesen. 2010. Can non-inversion tillage and straw retainment reduce N leaching in cereal-based crop rotations? Soil and Tillage Research 109(1):1-8. https://doi.org/10.1016/j.still.2010.04.001

Hansen, E. M., L. J. Munkholm, J. E. Olesen, and B. Melander. 2015. Nitrate leaching, yields and carbon sequestration after noninversion tillage, catch crops, and straw retention. Journal of Environmental Quality 44(3):868-881. https://doi.org/10.2134/jeq2014.11.0482

Hansen, E. M. Ø., J. Egmose, S. Lund, and H. Hauggaard-Nielsen. 2020. The role of experience exchange groups in farmers’ practice transitions: the case of reduced tillage in Danish conventional farming. Agroecology and Sustainable Food Systems 44(8):1012-1032. https://doi.org/10.1080/21683565.2020.1726550

Hastrup, F., N. Brichet, and L. R. Nielsen. 2022. Sustainable animal production in Denmark: anthropological interventions. Sustainability 14(9):5584. https://doi.org/10.3390/su14095584

Hazard, L., J. Locqueville, and F. Rey. 2022. A facilitation method to foster collective action in transitions toward sustainable agriculture—a case study. Agronomy for Sustainable Development 42(6):106. https://doi.org/10.1007/s13593-022-00838-9

Heath, S., A. Fuller, and B. Johnston. 2009. Chasing shadows: defining network boundaries in qualitative social network analysis. Qualitative Research 9(5):645-661. https://doi.org/10.1177/1468794109343631

Heleba, D., V. Grubinger, and H. Darby. 2016. On the ground. Putting agroecology to work through applied research and extension in Vermont. Pages 177-192 in V. E. Méndez, C. M. Bacon, R. Cohen, and S. R. Gliessman, editors. Agroecology: a transdisciplinary, participatory and action-oriented approach. CRC Press, Boca Raton, Florida, USA. https://doi.org/10.1201/b19500-12

Hendriks, C. M., and J. Grin. 2007. Contextualizing reflexive governance: the politics of Dutch transitions to sustainability. Journal of Environmental Policy & Planning 9(3-4):333-350. https://doi.org/10.1080/15239080701622790

Hoffecker, E. 2021. Understanding inclusive innovation processes in agricultural systems: a middle-range conceptual model. World Development 140:105382. https://doi.org/10.1016/j.worlddev.2020.105382

Hollstein, B. 2014. Qualitative approaches. Pages 404-416 in J. Scott and P. J. Carrington, editors. The SAGE Handbook of Social Network Analysis. SAGE Publications, London, UK. https://doi.org/10.4135/9781446294413.n27

Holmén, J., S. Williams, and J. Holmberg. 2022. Comparing sustainability transition labs across process, effects and impacts: insights from Canada and Sweden. Energy Research and Social Science 89. https://doi.org/10.1016/j.erss.2022.102522

Hurley, P. D., D. C. Rose, P. J. Burgess, and J. T. Staley. 2023. Barriers and enablers to uptake of agroecological and regenerative practices, and stakeholder views towards ‘living labs’. Page 33 in P. Burgess, J. Staley, P. D. Hurley, D. C. Rose, J. Redhead, M. E. McCracken, N. Girkin, L. Deeks, and J. Harris, editors. Evaluating the productivity, environmental sustainability and wider impacts of agroecological compared to conventional farming systems. Project SCF0321, Department for Environment, Food and Rural Affairs (DEFRA), Cranfield University, Cranfield, UK, and UK Centre for Ecology and Hydrology, Wallingford, UK.

International Energy Agency (IEA). 2023. Denmark climate resilience policy indicator. IEA, Paris, France. https://www.iea.org/reports/denmark-climate-resilience-policy-indicator

International Forum for Agroecology. 2015. Declaration of the International Forum for Agroecology, Nyéléni, Mali: 27 February 2015. Development 58(2-3):163-168. https://doi.org/10.1057/s41301-016-0014-4

Jarrahi, M. H., and S. Sawyer. 2019. Networks of innovation: the sociotechnical assemblage of tabletop computing. Research Policy 48:100001. https://doi.org/10.1016/j.repolx.2018.100001

Knierim, A., K. Boenning, M. Caggiano, A. Cristóvão, V. Dirimanova, T. Koehnen, P. Labarthe, and K. Prager. 2015. The AKIS concept and its relevance in selected EU member states. Outlook on Agriculture 44(1):29-36. https://doi.org/10.5367/oa.2015.0194

Koole, B. 2022. Veganism and plant-based protein crops: contentious visioning almost obstructing a transition. Environmental Innovation and Societal Transitions 42:88-98. https://doi.org/10.1016/j.eist.2021.12.003

Krafft, J., J. Höckert, M. Ljung, S. Lundberg, and C. L. Kolstrup. 2022. Delivering too much, too little or off target—possible consequences of differences in perceptions on agricultural advisory services. Agriculture and Human Values 39(1):185-199. https://doi.org/10.1007/s10460-021-10239-5

Lacoste, M., S. Cook, M. McNee, D. Gale, J. Ingram, V. Bellon-Maurel, T. MacMillan, R. Sylvester-Bradley, D. Kindred, R. Bramley, N. Tremblay, L. Longchamps, L. Thompson, J. Ruiz, F. O. García, B. Maxwell, T. Griffin, T. Oberthür, C. Huyghe, W. Zhang, J. McNamara, and A. Hall. 2022. On-farm experimentation to transform global agriculture. Nature Food 3(1):11-18. https://doi.org/10.1038/s43016-021-00424-4

Landini, F., F. Gallardo Lopez, G. Ramirez Valverde, M. Aguilar Calegare, and D. Kellen Onofre Dos Santos. 2021. How much do extension agents and advisors value agroecology in different countries? Contributions to the understanding of their potential role in scaling up agroecology. Agroecology and Sustainable Food Systems 45(10):1554-1578. https://doi.org/10.1080/21683565.2021.1933296

Leclère, M., L. Gorissen, Y. Cuijpers, L. Colombo, M. Schoonhoven-Speijer, and W. A. H. Rossing. 2024. Fostering action perspectives to support crop diversification: lessons from 25 change-oriented case studies across Europe. Agricultural Systems 218:103985. https://doi.org/10.1016/j.agsy.2024.103985

Leeuwis, C., L. Klerkx, and M. Schut. 2018. Reforming the research policy and impact culture in the CGIAR: integrating science and systemic capacity development. Global Food Security 16:17-21. https://doi.org/10.1016/j.gfs.2017.06.002

Levidow, L., M. Pimbert, and G. Vanloqueren. 2014. Agroecological research: conforming—or transforming the dominant agro-food regime? Agroecology and Sustainable Food Systems 38(10):1127-1155. https://doi.org/10.1080/21683565.2014.951459

Lohrum, N., M. Graversgaard, and C. Kjeldsen. 2021. Historical transition of a farming system towards industrialization: a Danish agricultural case study comparing sustainability in the 1840s and 2019. Sustainability 13(22):12926. https://doi.org/10.3390/su132212926

López-García, D., M. Cuéllar-Padilla, A. de Azevedo Olival, N. P. Laranjeira, V. E. Méndez, S. Peredo y Parada, C. A. Barbosa, C. Barrera Salas, M. Caswell, R. Cohen, A. Correro-Humanes, V. García-García, S. R. Gliessman, A. Pomar-León, A. Sastre-Morató, and G. Tendero-Acín. 2021. Building agroecology with people. Challenges of participatory methods to deepen on the agroecological transition in different contexts. Journal of Rural Studies 83:257-267. https://doi.org/10.1016/j.jrurstud.2021.02.003

López-Ridaura, S., O. Masera, and M. Astier. 2002. Evaluating the sustainability of complex socio-environmental systems. The MESMIS framework. Ecological Indicators 2(1-2):135-148. https://doi.org/10.1016/S1470-160X(02)00043-2

Mabry, L. 2008. Case study in social research. Pages 214-227 in P. Alasuutari, L. Bickman, and J. Brannen, editors. The SAGE handbook of social research methods. SAGE Publications, London, UK. https://doi.org/10.4135/9781446212165.n13

Magrini, M.-B., G. Martin, M.-A. Magne, M. Duru, N. Couix, L. Hazard, and G. Plumecocq. 2019. Agroecological transition from farms to territorialised agri-food systems: issues and drivers. Pages 69-98 in J. E. Bergez, E. Audouin, and O. Therond, editors. Agroecological transitions: from theory to practice in local participatory design. Springer, Cham, Switzerland. https://doi.org/10.1007/978-3-030-01953-2_5

Meadows, D. 1999. Leverage points: places to intervene in a system. The Sustainability Institute, Hartland, Wisconsin, USA.

Méndez, V., C. M. Bacon, R. Cohen, and S. R. Gliessman. 2016. Agroecology: a transdisciplinary, participatory and action-oriented approach. CRC Press, Boca Raton, USA. https://doi.org/10.1201/b19500

Méndez, V., M. Caswell, S. Gliessman, and R. Cohen. 2017. Integrating agroecology and participatory action research (PAR): lessons from Central America. Sustainability 9(5):705. https://doi.org/10.3390/su9050705

Meng, D., M. Du, and Y. Wu. 2022. Overview of signed networks. Pages 1-13 in D. Meng, M. Du, and Y. Wu, editors. Disagreement behavior analysis of signed networks. Springer Nature, Singapore. https://doi.org/10.1007/978-981-19-5530-3_1

Mier y Terán Giménez Cacho, M., O. F. Giraldo, M. Aldasoro, H. Morales, B. G. Ferguson, P. Rosset, A. Khadse, and C. Campos. 2018. Bringing agroecology to scale: key drivers and emblematic cases. Agroecology and Sustainable Food Systems 42(6):637-665. https://doi.org/10.1080/21683565.2018.1443313

Millenium Ecosystem Assessment. 2005. Ecosystems and human well-being: a framework for assessment. Island Press, Washington, D.C., USA. https://www.millenniumassessment.org/documents/document.356.aspx.pdf

Ministry of Food, Agriculture and Fisheries. 2021. The agreement on a green transition of the agricultural sector. Ministry of Food, Agriculture and Fisheries of Denmark, Copenhagen, Denmark. https://en.fvm.dk/news-and-contact/focus-on/the-agreement-on-a-green-transition-of-the-agricultural-sector

Moraine, M., S. Lumbroso, and X. Poux. 2016. A comprehensive outlook on the diversity of agroecological initiatives in Europe—strategies for contrasting crop diversification strategies in Europe. Institute for Sustainable Development and International Relations (IDDRI), Paris, France. https://doi.org/10.13140/RG.2.1.1546.7121

Morel, K., E. Revoyron, M. San Cristobal, and P. V. Baret. 2020. Innovating within or outside dominant food systems? Different challenges for contrasting crop diversification strategies in Europe. PLoS ONE 15(3):e0229910. https://doi.org/10.1371/journal.pone.0229910

Mottet, A., A. Bicksler, D. Lucantoni, F. De Rosa, B. Scherf, E. Scopel, S. López-Ridaura, B. Gemmil-Herren, R. Bezner Kerr, J.-M. Sourisseau, P. Petersen, J.-L. Chotte, A. Loconto, and P. Tittonell. 2020. Assessing transitions to sustainable agricultural and food systems: a tool for agroecology performance evaluation (TAPE). Frontiers in Sustainable Food Systems 4:579154. https://doi.org/10.3389/fsufs.2020.579154

Mozzato, D., P. Gatto, E. Defrancesco, L. Bortolini, F. Pirotti, E. Pisani, and L. Sartori. 2018. The role of factors affecting the adoption of environmentally friendly farming practices: can geographical context and time explain the differences emerging from literature? Sustainability (Switzerland) 10(9):3101. https://doi.org/10.3390/su10093101

Munkhom, L. J., E. M. Hansen, B. Melander, P. Kudsk, L. N. Jørgensen, G. J. Heckrath, S. Ravnskov, and J. A. Axelsen. 2020. Vidensyntese om conservation agriculture. Aarhus Universitet, DCA - Nationalt Center for Fødevarer og Jordbrug, Tjele, Denmark.

Nesheim, M. C., M. Oria, and P. T. Yih, editors. 2015. Overview of the U.S. food system. In Institute of Medicine and National Research Council. A framework for assessing effects of the food system. National Academies Press, Washington, D.C., USA. https://doi.org/10.17226/18846

Norris, P. E., M. O’Rourke, A. S. Mayer, and K. E. Halvorsen. 2016. Managing the wicked problem of transdisciplinary team formation in socio-ecological systems. Landscape and Urban Planning 154:115-122. https://doi.org/10.1016/j.landurbplan.2016.01.008

Opsahl, T., F. Agneessens, and J. Skvoretz. 2010. Node centrality in weighted networks: generalizing degree and shortest paths. Social Networks 32(3):245-251. https://doi.org/10.1016/j.socnet.2010.03.006

Organisation for Economic Co-operation and Development (OECD). 2019a. Income inequality. OECD, Paris, France.

OECD. 2019b. Poverty rate. OECD, Paris, France.

OECD. 2021. Gross domestic product (GDP). OECD, Paris, France.

Osei-Owusu, A. K., R. Wood, E. L. Bjelle, D. Caro, and M. Thomsen. 2021. Understanding the trends in Denmark’s global food trade-related greenhouse gas and resource footprint. Journal of Cleaner Production 313:127785. https://doi.org/10.1016/j.jclepro.2021.127785

Pedersen, H. B., L. M. Arendt, C. F. Slothuus, and F. S. Møller. 2022. Portræt af danske landmænd. Danmarks Statistik, Copenhagen, Denmark.

Polge, E., and H. Pagès. 2022. Relational drivers of the agroecological transition: an analysis of farmer trajectories in the Limagne plain, France. Agricultural Systems 200:103430. https://doi.org/10.1016/j.agsy.2022.103430

Reipurth, M. F. S., L. Hørby, C. G. Gregersen, A. Bonke, and F. J. A. Perez Cueto. 2019. Barriers and facilitators towards adopting a more plant-based diet in a sample of Danish consumers. Food Quality and Preference 73:288-292. https://doi.org/10.1016/j.foodqual.2018.10.012

Rocker, S., J. Kropczynski, and C. Hinrichs. 2022. Using social network analysis to understand and enhance local and regional food systems. Pages 231-256 in C. J. Peters and D. D. Thilmany, editors. Food systems modelling. Academic Press, Elsevier, London, UK. https://doi.org/10.1016/B978-0-12-822112-9.00015-1

Rodenburg, J., L. Büchi, and J. Haggar. 2020. Adoption by adaptation: moving from conservation agriculture to conservation practices. International Journal of Agricultural Sustainability 19(5-6):437-455. https://doi.org/10.1080/14735903.2020.1785734

Rossing, W. A. H., M. M. Albicette, V. Aguerre, C. Leoni, A. Ruggia, and S. Dogliotti. 2021. Crafting actionable knowledge on ecological intensification: lessons from co-innovation approaches in Uruguay and Europe. Agricultural Systems 190:103103. https://doi.org/10.1016/j.agsy.2021.103103

Rossing, W. A. H., S. Dogliotti, and G. Martin. 2023. Sustainability transitions in the making in agroecosystems: changes in research scope and methods. Agricultural Systems 210:103719. https://doi.org/10.1016/j.agsy.2023.103719

Ryschawy, J., S. Tiffany, A. Gaudin, M. T. Niles, and R. D. Garrett. 2021. Moving niche agroecological initiatives to the mainstream: a case-study of sheep–vineyard integration in California. Land Use Policy 109:105680. https://doi.org/10.1016/j.landusepol.2021.105680

Schoonhoven, Y., and H. Runhaar. 2018. Conditions for the adoption of agro-ecological farming practices: a holistic framework illustrated with the case of almond farming in Andalusia. International Journal of Agricultural Sustainability 16(6):442-454. https://doi.org/10.1080/14735903.2018.1537664

Schurman, R. 2018. Micro(soft) managing a ‘green revolution’ for Africa: The new donor culture and international agricultural development. World Development 112:180-192. https://doi.org/10.1016/j.worlddev.2018.08.003

shiftN. 2023. Global food system map. shiftN, Leuven, The Netherlands. https://shiftn.com/_uploads_pdf/shiftN-Global-Food-System-Maps-kopie.pdf

StatBank Denmark. n.d. Cultivated area by unit, region, crop and time. Statistics Denmark, Copenhagen, Denmark. https://www.statbank.dk/AFG5

StatBank Denmark. 2023. BDF11: farms by region, unit, type of farms and area. Statistics Denmark, Copenhagen, Denmark. https://www.dst.dk/en

Statistics Denmark. n.d.a. Cropland. Statistics Denmark, Copenhagen, Denmark. https://www.dst.dk/en/Statistik/emner/erhvervsliv/landbrug-gartneri-og-skovbrug/det-dyrkede-areal

Statistics Denmark. n.d.b. Livestock. Statistics Denmark, Copenhagen, Denmark. https://www.dst.dk/en/Statistik/emner/erhvervsliv/landbrug-gartneri-og-skovbrug/bestanden-af-husdyr

Statistics Denmark. 2018. Kun få unge landmænd i Danmark. Statistics Denmark, Copenhagen, Denmark. https://www.dst.dk/da/Statistik/nyheder-analyser-publ/nyt/NytHtml?cid=25859

Statistics Denmark. 2021. Land use accounts. Statistics Denmark, Copenhagen, Denmark. https://www.dst.dk/en/Statistik/emner/miljoe-og-energi/areal/arealopgoerelser

Statistics Denmark. 2024. NYT: Markant fald i antal bedrifter. Statistics Denmark, Copenhagen, Denmark. https://www.dst.dk/da/Statistik/nyheder-analyser-publ/nyt/NytHtml?cid=38302

United Nations Development Programme (UNDP). 2022. Human development report 2021/2022. United Nations Development Programme, New York, New York, USA.

van Mierlo, B., A. Janssen, F. Leenstra, and E. van Weeghel. 2013. Encouraging system learning in two poultry subsectors. Agricultural Systems 115:29-40. https://doi.org/10.1016/j.agsy.2012.10.002

van Mierlo, B., B. Regeer, M. van Amstel, M. C. M. Arkesteijn, V. Beekman, J. F. G. Bunders, T. de Cock Buning, B. Elzen, A. C. Hoes, and C. Leeuwis. 2010. Reflexive monitoring in action. A guide for monitoring system innovation projects. Wageningen University and Research, Wageningen/Amsterdam, The Netherlands.

Vanwindekens, F. M., P. V. Baret, and D. Stilmant. 2014. A new approach for comparing and categorizing farmers’ systems of practice based on cognitive mapping and graph theory indicators. Ecological Modelling 274:1-11. https://doi.org/10.1016/j.ecolmodel.2013.11.026

Vanwindekens, F. M., D. Stilmant, and P. V. Baret. 2013. Development of a broadened cognitive mapping approach for analysing systems of practices in social-ecological systems. Ecological Modelling 250:352-362. https://doi.org/10.1016/j.ecolmodel.2012.11.023

Wezel, A., B. G. Herren, R. B. Kerr, E. Barrios, A. L. R. Gonçalves, and F. Sinclair. 2020. Agroecological principles and elements and their implications for transitioning to sustainable food systems. A review. Agronomy for Sustainable Development 40(6):40. https://doi.org/10.1007/s13593-020-00646-z

Willems, J., H. J. M. van Grinsven, B. H. Jacobsen, T. Jensen, T. Dalgaard, H. Westhoek, and I. S. Kristensen. 2016. Why Danish pig farms have far more land and pigs than Dutch farms? Implications for feed supply, manure recycling and production costs. Agricultural Systems 144:122-132. https://doi.org/10.1016/j.agsy.2016.02.002

World Topographic Map. n.d. Denmark topographic map. https://en-gb.topographic-map.com/map-z61h/Denmark/

Xiang, W.-N. 2013. Working with wicked problems in socio-ecological systems: awareness, acceptance, and adaptation. Landscape and Urban Planning 110:1-4. https://doi.org/10.1016/j.landurbplan.2012.11.006

Yousefi Nooraie, R., J. E. M. Sale, A. Marin, and L. E. Ross. 2020. Social network analysis: an example of fusion between quantitative and qualitative methods. Journal of Mixed Methods Research 14(1):110-124. https://doi.org/10.1177/1558689818804060

Zurek, M., J. Wirths, A. Hebinck, S. Crawford, P. Lidder, P. V. V. Prasad, P. Tittonell, M. Herrero, and J. Compton. 2023. Principles for guiding research and innovation toward sustainable and equitable agrifood systems. Frontiers in Sustainable Food Systems 7:1059063. https://doi.org/10.3389/fsufs.2023.1059063