Navigation – Plan du site

AccueilNuméros25-2From territorial diagnosis to str...

From territorial diagnosis to strategic intent. Knowledge brokerage and co-construction. Fieldwork Insights from Karst Landscapes – the poljes of Canton 10 (Bosnia and Herzegovina)

Orianne Crouteix, François Lerin et Jérémy Vendé
Cet article est une traduction de :
Du diagnostic territorial à l’intention stratégique. Courtage de connaissances et coconstruction. Retour de terrains en pays karstique – les poljés du Canton 10 (Bosnie et Herzégovine) [fr]

Résumés

Cet article défend l’idée qu’un diagnostic territorial, dans une démarche inductive et compréhensive, permet de développer une approche intégrée et systémique, et ainsi de s’extraire des logiques thématiques (en silo) qui contraignent l’action environnementale. Notre diagnostic du territoire karstique du Canton 10 de Bosnie-et-Herzégovine combine trois cadres d'analyse issus des sciences de gestion et de la géographie : l'analyse stratégique de la gestion environnementale, un diagnostic agraire et des approches paysagères. Nous assumons une intentionnalité environnementale sur et pour le territoire. Avec nos bricolages méthodologiques et notre travail itératif, nous insistons sur (i) l’importance et le caractère remarquable et unique d’un système hydro-géomorphologique, (ii) la nécessité d’enrôler les acteurs agricoles et de comprendre les dynamiques liées au système d’élevages. En effet, notre diagnostic permet de mettre en évidence des dynamiques territoriales cruciales qui sont souvent négligées par les projets environnementaux descendants, telles que la présence d'un pastoralisme extensif, les fortes disparités d'habitat liées aux enjeux démographiques, politiques et historiques du territoire et le développement des énergies dites vertes. Paradoxalement, et à rebours des projets de conservation portés par des fondations internationales qui se concentrent sur un unique poljé (une grande dépression karstique à fond plat) avec une reconnaissance nationale, nous suggérons un levier d’action au niveau du canton avec une reconnaissance internationale. On discute alors de l’importance et des limites de ces bricolages méthodologiques, du courtage de connaissances et de cette posture embarquée.

Haut de page

Texte intégral

The authors wish to thank the 2022–2023 cohort of the FNS-MI specialized master’s program, as well as Emmanuel Artus and Anouk Fraisse, for their fieldwork and the numerous exchanges held with the authors. This research, and particularly the field investigation, was supported by AgroParisTech. The authors also thank the reviewers for their comments, which significantly enriched and improved this article.

Introduction

1This article presents feedback from fieldworks carried out in a region of southern Bosnia and Herzegovina (B&H). It outlines the steps involved in the gradual development, with stakeholders, of territorial diagnoses for an agri-environmental initiative. It also reports on our role as brokers, both in terms of knowledge and in the development of a co-constructed project with local actors. It is a reflection on an experience from an environmental diagnosis toward the development of a shared territorial project.

2The territory was selected for its remarkable environmental and landscape qualities, as well as for the importance of its forest and pastoral resources and activities. These play a key role in sustaining meat and cheese production across the wider region. The karst polje region of Canton 10 in the Federation of Bosnia and Herzegovina constitutes a spectacular and complex geomorphological and hydrological system in the western Dinaric Alps (within the geomorphological subdivision of the Western Bosnian Mountains and the Dinara Massif). In the local language, polje means “plain,” but the term has been adopted in geographical vocabulary to designate: “a closed depression of varying size (from a few kilometres to several dozen kilometres) in a karst landscape; its origin is often tectonic; the underground drainage of water prevents ‘normal’ erosion and runoff; during heavy rainfall, the lowest parts may become temporarily flooded” (Brunet et al., 2005, p. 394). These limestone formations are also considered fragile and threatened by human activities (Brinkmann & Parise, 2012).

3Contact with the study area first occurred in the context of a comparative diagnostic aimed at identifying environmental issues in two municipalities, Sanski Most and Mrkonjić Grad (Malešević & Siegel, 2023). The latter municipality includes a Mediterranean causse landscape that contrasts with the predominantly mountainous areas in the study region. Our attention was drawn to this landscape as a precursor to the more southern karst system. At the same time, the Bosnian non-governmental organization (NGO) CZZS (Centar za životnu sredinu, Center for the Environment1), with which we had carried out this comparative diagnostic, was conducting a biodiversity project in the major poljes located in Canton 10 of the Federation of Bosnia and Herzegovina. These karst landscapes are subject to multiple initiatives: the creation of a protected area on part of the territory, the promotion of agricultural products through geographical indications (GI), and energy infrastructure projects. However, these initiatives often struggle to account for the diversity of uses and representations, highlighting the absence of a shared territorial diagnostic that integrates ecological, economic, social, and cultural dimensions of the landscape. We therefore set out to develop a collective diagnostic for the area of these large poljes.

4Based on the first data available and our discussions with CZZS, we framed this diagnostic around three questions: (1) How can the territory be delineated so that it makes sense from a physical perspective while also accounting for the human shaping of its landscapes and management units? (2) Within this territory, how can environmental and economic issues be characterized, ranked, and interconnected? (3) In light of this first diagnostic, what forms of protection would be most effective for safeguarding and enhancing these natural and cultural heritages?

5In this context, the aim of this article is to present a territorial diagnostic approach based on a deliberate methodological bricolage, combining agrarian diagnostics, strategic environment management analysis, and landscape analysis. Our intention is to build a tool of strategic intentionality, designed to shed light on the tensions and potentials of the territory through a co-constructive approach. With this article, we seek to contribute to ongoing reflections on how territorial diagnostics are crafted in complex, multi-actor contexts, highlighting the need for inductive and reflexive approaches to support the agri-environmental transition of rural areas.

6The first three sections of this article are devoted to our approach: the first outlines our research posture, drawing on embedded research, knowledge brokering, and methodological bricolage; the second specifies the analytical frameworks mobilized to co-construct the territorial diagnostic; and the third sets out the different phases in the diagnostic process. The fourth section presents the results obtained by following this inductive approach. The fifth discusses the structuring of a learning territory and the role of knowledge brokering.

Research Stance and Methodological Approach

Embedded Research: Theoretical Contributions

7The approach adopted in this work is grounded in an embedded research posture, as defined by Bernard-Mongin (2016), which draws extensively on the literature on post-normal science (Funtowicz & Ravetz, 1993) and subsequent developments. This embedded research framework makes it possible “to simultaneously hold together the exploratory and intentional dimensions of constructing an environmental action in context” (Bernard-Mongin, 2016, p. 16). “Similar to action research, embedded research generates its material from interactions with the field but goes further in that its inquiry is motivated and structured by an environmental intention” (Bernard-Mongin, 2016, p. 19). This posture does not simply entail the researcher’s personal involvement during the fieldwork; rather, it signifies a profound transformation of the researcher’s role, responsibilities, and modes of interaction in the production of knowledge (Nogueira et al., 2021).

8The researcher is neither an external observer nor a commissioned expert. Instead, they position themselves as one actor among others, tasked with opening spaces for dialogue—or “hybrid forums”—and translating heterogeneous forms of knowledge (Callon et al., 2001). In this regard, the researcher embraces their own ignorance, and it is through reflection and discussion that knowledge is generated (Les chercheurs ignorants, 2015). This position also resonates with the principles of Actor–Network Theory (Callon, 1986; Akrich et al., 2006; Guinchard et al., 2023) and organizational sociology (Crozier & Friedberg, 1977), both of which emphasize that knowledge production is constructed through the enrolment, mobilization, and integration of actors within unstable networks. Relationships between researchers and local stakeholders, their respective commitments, the tools mobilized, and the conditions of access to the field all influence what is—or will be—observed, produced, and legitimized.

9Embedded research, which concretely shapes methodological choices, the tools deployed, and the mode of dissemination is, by definition, dynamic and progressive with and for the territory. This dynamism manifests both in the shifting position of researchers in relation to local actors and in the methodologies and forms of knowledge that must be mobilized. There are thus no three distinct stages—formulation of theory, application, and dissemination—but rather constant back-and-forth movements between data collection, analysis, restitution, and the formulation of proposals.

Knowledge Brokerage: Foundations and Practices

10The aim of this work is to understand territorial dynamics to co-produce actionable knowledge. Co-construction is not necessarily synonymous with consensus around a single solution. The analysis of local dynamics must identify both consensus and dissent, as well as potential levers for action. In this sense, researchers become knowledge brokers and “question framers”, shifting perspectives and proposing new visions to co-create new territorial narratives. The notion of knowledge brokerage, as mobilized here, designates an intermediate posture between scientific production, socio-political mediation, and territorial facilitation. In this position, the researcher is neither a mere expert, nor a facilitator, nor a decision engineer: they act as an intermediary, a translator of knowledge between distinct social, cultural, and cognitive worlds.

11In contexts marked by complexity, uncertainty, and conflicts of legitimacy, this role is central to enabling the circulation of ideas, making asymmetries visible, and co-constructing pathways for action—particularly when the aim is to generate strategic intentionality, as we will illustrate below.

12This positioning is directly inspired by work on “technical democracy” (Callon et al., 2001), which calls for science to be taken outside expert enclosures to better connect it to secular, situated, and experiential forms of knowledge. Within this theoretical framework, the researcher does not deliver a solution or a definitive truth but instead organizes the conditions for an informed, reflective, and evolving debate. Knowledge brokerage is thus not about simplifying complexity but rather about collectively deepening it, by translating analytical categories, constraints, expectations, and horizons of action across groups of stakeholders.

13The knowledge broker is therefore not neutral but must maintain a stance of “productive misalignment,” enabling them to shed light on blind spots, reframe unspoken assumptions, and introduce cross-cutting perspectives where sectoral approaches dominate.

14Key notions such as enrolment, controversy, and the negotiation of meanings are thus crucial to understand how strategic intentionality is formed within our framework. In this study, we demonstrate how this brokerage dimension materialized through translations across scales of analysis (from the polje to the municipality, from the canton to the European level), through the dialogue between different forms of knowledge (farmers, technicians, municipalities, researchers, NGOs), and through the progressive enrolment of heterogeneous actors around shared territorial hypotheses.

Methodological Bricolage as a Strategy of Deliberate Intentionality

15The analytical premise underpinning this work is that inquiry should not begin from a pre-determined entry point—be it biodiversity, agriculture, or rewilding—thereby risking being constrained or even trapped within a siloed approach. Instead, we advocate for an inductive process and a comprehensive understanding of the territory, without any a priori definition or delimitation of either the territory itself or the environmental and development issues at stake. While the identification of the geographical area and of environmental challenges are prerequisites for territorial projects, they cannot be taken for granted; they must be defined and discussed (Gasnier et al., 2022).

16Such a territorial approach appears to us to be the most appropriate for prioritizing the issues at the climate change / biodiversity loss nexus. This requires, on the one hand, an investigation into the dynamics of local actors and the rationale for their production and management choices, and on the other, an understanding of how environmental issues are identified and framed. Overcoming siloed perspectives and producing co-constructed knowledge for environmental action, while critically assessing the potential biases of data produced by single-discipline research, calls for methodological dynamism and reflexivity. This is what we term bricolage—an adaptive, context-sensitive posture that responds to the specific challenges of the actors and territories involved.

17The term methodological bricolage, often associated with improvisation or epistemological weakness, is here reclaimed as a deliberate strategy of adaptation to the complexity of reality, in line with the post-normal science framework (Funtowicz & Ravetz, 1993) and the work of Gibbons et al. (1994). In situations marked by uncertainty and by a plurality of interests (some even conflicting), methodological rigidity is more an obstacle than an asset. Bricolage, conversely, enables the emergence of inductive intelligence, combining observation, experimentation, iteration, and interpretation. It thus allows us to “solve problems or attempt to answer questions that have hitherto remained unanswered” (Raffestin, 2003, p. 71).

18The form of methodological bricolage deployed here is both heuristic and institutional. It is heuristic insofar as it requires constant reflexivity regarding the choices made and the tools selected, and because it creates bridges between different registers of knowledge—scientific, technical, and local—depending on the issues encountered. It is institutional in that it entails engaging with academic formats that are attuned to the field and responsive to the expectations of local stakeholders, as well as to the logics of territorialized projects (cross-border cooperation, rural development, heritage designation, etc.).

19This bricolage also calls for a reversal of the conventional temporalities of research. In such a posture, the goal is not to apply a fixed methodology or a single analytical framework but to mobilize a plurality of tools and concepts that respond to the specificities of the fieldwork situation. In contrast to a strict disciplinary methodology, which begins with a pre-formulated research question and expects the field to deliver its answer, our approach asks which data, tools, concepts, and categories can be most useful for understanding the territorial issues at stake, for structuring the inquiry process, and for co-constructing subsequent steps with the relevant actors. This bricolage “finds its methodological justification and heuristic significance in the social sciences in that it authorizes and values the construction of composite research designs closely related to the object of study, allowing for a sharper rendering of the object’s concrete reality” (Feildel, 2016, p. 97–98). The methodological plan is thus not entirely fixed ex ante but is adjusted progressively as encounters, resistances, and investigative opportunities arise. Rather than proposing a default methodology, we advocate for a pragmatic bricolage, whose effectiveness is measured by its capacity to foster dialogue between disciplines and experiences, while producing robust, shared, and action-relevant forms of knowledge for territorial governance.

20This bricolage develops iteratively over the course of the inquiry, shaped by opportunities and by available funding mechanisms that enable at least partial responses to the research questions and thereby allow the investigation to continue. While this approach frees us from dependence on a single project or funding source—each with its own deadlines and disbursement conditions—it requires continuous monitoring of funding opportunities and an ability to adapt to the constraints of each project and partnership. Combining funding streams and projects allows us to embed territorial approaches over time and to integrate a diversity of thematic perspectives. However, this also demands organizational, methodological, and administrative flexibility, which can be time- and resource-intensive.

Analytical Frameworks Mobilized

Stated Objectives

21Our diagnostic approach was designed with two main objectives. First, to break free from the environmental silos generated by project-based funding schemes, international databases, or narrowly disciplinary approaches. Second, to produce data, information, and knowledge that could serve as a foundation for actionable, territorially embedded environmental interventions. Such knowledge must be contextually relevant, with its accuracy and precision co-constructed in collaboration with stakeholders—most of whom are outside the academic sphere.

22To achieve these objectives, it is necessary to adopt an inductive, territory-based, and comprehensive approach. This approach takes a critical stance toward available environmental information, interrogating discourses, typologies, and data to identify implicit assumptions. Through this process of deconstruction, these elements acquire a consolidated epistemic status, supported by reflexivity (in other words: what is their level of relevance, for which type of analysis or diagnosis, and within what argumentative or descriptive framework?). For instance, one must critically examine top-down approaches that emphasize the importance of migratory birds, wetlands, or deforestation when such emphasis derives more from the mandate of the issuing organization, the nature of the data mobilized, or the availability of funding, rather than from a genuinely territorialized assessment of environmental priorities.

23In this sense, it would be inaccurate to call these territories “case studies” or “field sites,” as both the territories and their actors are deeply involved in the knowledge production process. We find it preferable to call them “learning territories” (Bernard et al., 2023). Our approach relies on the progressive assemblage of knowledge, concepts, tools, and methods from various disciplinary traditions: (i) those oriented toward describing environments (geography, ecology, hydrology, soil science, agronomy); (ii) those devoted to analysing social issues (socio-anthropological methods, human geography); and (iii) those designed for action (management sciences). This methodological weaving is not linear but follows a logic of mutual adjustment between emergent field questions, actors’ modes of engagement, and the resources that can be mobilized at each stage.

24Two key features of our stance must be highlighted—albeit in a necessarily partial way. First, an explicit environmental intentionality motivated by the urgency of contemporary ecological challenges (climate change, biodiversity loss, resource depletion and pollution). Second, an “agri-environmental” bias or lens, grounded in the belief that agricultural and forestry practices are crucial for understanding territorial transition strategies, particularly in rural areas. These two considerations partly explain our choice of this specific region of Bosnia and Herzegovina: a fragile yet exceptional karstic area, notable for its hydro-geomorphological functioning, its landscapes, and its agri-silvo-pastoral history.

Methodological Composition of the Diagnostic

25Our territorial diagnostic relies on three principal analytical and methodological pillars: (1) Strategic Environment Management Analysis (SEMA); (2) Agrarian Diagnosis; (3) Landscape Approaches developed by geographers and agronomists within the francophone tradition.

26The purpose of combining these three frameworks is to analyse the territory, with explicit environmental intentionality, in all its complexity, integrating spatial organizations, actor systems, and their dynamics (Moine, 2006).

27SEMA is a theoretical framework developed within management sciences and tested on various environmental cases in France and abroad since the late 1980s (Mermet, 2014; Mermet et al., 2005). It rests on three axioms: that when confronted with a complex situation, interpretation must adopt a partial and explicitly problem-centred perspective; that the analysis of actor strategies must be informed by organizational sociology (Crozier & Friedberg, 1977); and that it is essential to cultivate a strategic perspective within environmental management. SEMA is structured around four key concepts:

  • Reference Environmental Concern or Normative Framework: the explicit definition of the environmental issue and its precise measurement criteria.

  • Effective Management: the description of the system, practices impacting the environment, and actor relationships.

  • Environmental Actors: those who hold and operationalize the reference concern, implementing practices and uses that support the achievement of the environmental objective—collectively referred to as intentional management.

  • Strategic Dimension: the set of actions to be implemented by environmental actors to conserve or restore the initially defined environmental concern, in the face of other actors’ objectives.

28An environmental diagnostic must therefore identify the “environmental concern” and initiate the description of “effective management”—the first two key concepts of SEMA. The latter can be described using organizational sociology and semi-structured interviews, common tools within the social sciences. However, defining the normative framework often requires reliance on work by ecologists, biologists, or other experts (academic, NGO, or consulting), with limited capacity to interrogate how such information is produced and used, and thus remains dependent on theme-specific, siloed approaches (wetland biodiversity, deforestation, pollution, etc.). To overcome these two limitations, we complement SEMA with two additional tools: agrarian diagnosis and landscape analysis.

29Agrarian diagnosis provides a systematic description of the production systems and agrarian systems of a territory or terroir (Cochet, 2011a; 2011b). Although the concept of agrarian system was first defined by geographers (Cholley, 1946), it has been further developed by economists and agronomists who established a methodological framework for the analysis and diagnosis of regions. The aim is to analyse the ongoing dynamics in each region and to reconstruct “the trajectories of evolution of the different production systems from an action-oriented perspective” (Cochet, 2011a, p. 91), while compiling the data required for economic or environmental evaluation, or for the design of an agricultural project or policy. Originally conceptualized by Mazoyer in the 1970s (building on Dumont’s teachings), this methodology has been continuously refined, notably by Dufumier, agronomists, and their field-based feedback (Cochet, 2011a).

30Agrarian diagnosis involves extensive fieldwork and is composed of six stages: 1. delimitation of the agrarian region, 2. landscape reading and analysis, 3. reconstruction of the history of agrarian structures through semi-structured interviews, 4. identification of typologies of agricultural production systems using the results of the previous steps combined with statistical and regional data analysis, 5. economic analysis through surveys of farmers and herders, 6. evaluation according to the objectives of the study (Cochet, 2011a).

31This approach yields a precise picture of agricultural production systems and their dynamics, which, in rural contexts, are often directly linked to environmental issues. Agricultural practices can either contribute to environmental degradation (e.g., nitrate pollution) or act as positive levers for addressing key environmental concerns (e.g., pastoralism as a tool against scrub encroachment). However, agrarian diagnosis can risk overemphasizing the agricultural dimension of the territory. For this reason, we complement it with a landscape-based approach.

32In geography, the concept of landscape has been extensively debated. At times, geographers have valorised the term to the point that “some have claimed that geography was the science of landscapes” (Brunet et al., 2005, p. 373), while others have sharply criticized it for “selecting certain elements and losing sight of the whole, whether in agrarian or morphological landscapes. Through what was seen and what was deliberately overlooked, the landscape sustained a science of fragmentary appearances and physical determinations” (Brunet et al., 2005, p. 375). Environmental geography, with its intent to integrate biophysical, spatial, and historical dimensions, readily mobilizes the concept of landscape as a nature–culture interface (Blanc-Pamard, 1986; Bertrand & Bertrand, 2002). Landscape analysis also serves as a bridge between geography and ecology (Demailly, 2017), particularly in its capacity to interrogate vegetation patterns in relation to anthropogenic dynamics (Ballouche, 2016). Some scholars even argue that “landscape ecology is a corpus of hypotheses and tools that provide an interdisciplinary exchange platform between the social sciences studying landscape and ecological approaches incorporating spatial variables” (Marty et al., 2006, p. 357).

33Landscape reading and analysis thus make it possible to conceptualize the interfaces between the physical and human components of the environment, the ecological and socio-economic dynamics, and to embrace the production of situated knowledge and a sensitive, territory-based approach (Davodeau, 2005; Joliveau et al., 2008).

34Together, these three analytical frameworks are action-oriented. SEMA ultimately seeks to propose strategies for environmental actors to respond to the reference concern. The final stage of agrarian diagnosis, formalized by agronomists within a rural development logic, constitutes an evaluation designed to address economic and environmental issues. Finally, landscape ecology approaches are explicitly intended as tools to co-design land-use and environmental protection projects with local stakeholders (Cottet, 2019; Décamps & Décamps, 2007; Nassauer, 1992). These three approaches were each developed within action-research traditions focused on environmental and rural development issues. By combining and adapting them, we explicitly embrace an environmental intentionality: the ultimate objective of our research (whether achieved or not) is to contribute to the construction and design of territorial environmental or agri-environmental projects.

Figure 1: Articulation of the three analytical frameworks used in this study (according to the authors)

Figure 1: Articulation of the three analytical frameworks used in this study (according to the authors)

Fieldwork and Data

Phase 1 – Identification of a Territory and Initial Partnerships (2022)

35In 2022, an initial study was conducted by a specialized master’s student from the Forêt, Nature et Société – Management International (MS FNS-MI) program at AgroParisTech, together with a geographer from the University of Banja Luka, in collaboration with the environmental NGO CZZS. The study focused on the municipalities of Mrkonjić Grad and Sanski Most (Malešević & Siegel, 2023) and was made possible thanks to funding from the European Climate Foundation (ECF). The work was supervised by two of the present article’s authors, and two field-support missions were organized. These missions also provided opportunities to discuss with our Bosnian partners their ongoing project in Canton 10, specifically on the Livanjsko polje, and more broadly on the three other major poljes in the Canton (Duvanjsko, Glamočko, Kupreško). This project is supported and partially implemented by the German foundation EuroNatur.

36Together with CZZS and EuroNatur, we agreed that it would be useful to complement their work with a more comprehensive territorial and environmental diagnostic analysis, while initiating a reflection on potential labels, statuses, and protection networks that might be envisaged for such a territory. During this phase, we also established contact with the French Embassy in Sarajevo and several universities in Bosnia and Herzegovina.

37This first phase thus made it possible to identify new partners, co-construct research questions with environmental NGOs, initiate bibliographic work, identify territorial experts, and gather preliminary data (such as maps and project or mission reports). The main data collection techniques in this phase ranged from participant observation (during meetings and online interviews) to bibliographic research and the gathering of grey literature.

Phase 2 – Collective Field Mission and Multi-Actor Diagnostic (2023)

38The fieldwork was carried out as part of the MS FNS-MI program at AgroParisTech, which includes the completion of a three-week international collective field mission. In February–March 2023, this mission was conducted in Livno (Bosnia and Herzegovina) and involved two faculty members from AgroParisTech, a researcher from the association AIDA2, and fourteen participants from the specialized master’s program.

39The mission pursued multiple objectives: pedagogical (for the students), scientific (knowledge production for the researchers), operational (data collection and network development for field partners), among others. This multiplicity of goals can be a constraint but also offers the advantage of collecting substantial data in a short period of time and combining the expertise and perspectives of all participants. With the support of four students from the University of Sarajevo, it was possible to conduct 59 semi-structured interviews with elected officials, economic actors (agriculture, tourism, energy), public environmental management agencies, and NGOs.

40The purpose of these interviews was to map the actor system (Figure 2), to identify and propose a hierarchy of environmental issues, and to more precisely delimit the study area, including the four major poljes. In parallel, landscape reading and analysis were conducted across each of the poljes. Additional data were gathered through the analysis of technical reports (from NGOs or public agencies) and cartographic resources.

41Finally, two restitution sessions (one in Livno and one in Sarajevo) allowed us to present preliminary findings to local and national stakeholders and refine them through discussion. This work also resulted in the drafting of a collective report detailing the methodology, results, and an initial reflection on potential protection statuses (Bricault et al., 2023).

Figure 2: Schematic representation of the actor system developed during this diagnostic phase (Bricault et al., 2023, p. 63)

Figure 2: Schematic representation of the actor system developed during this diagnostic phase (Bricault et al., 2023, p. 63)

Phase 3 – Agrarian Diagnostic and Sectoral Deepening (2024)

42This third phase made it possible to clarify two aspects that required further investigation: (i) the different livestock systems and their uses in each landscape unit (particularly in relation to the maintenance of open habitats), and (ii) the role of the geographical indication (GI) for Livno cheese in shaping agropastoral dynamics, particularly regarding water management (both in terms of quantity and quality), identified as a major environmental issue in this karst region.

43An agrarian diagnostic was carried out between March and November 2024 by two students from the Institut Agro, with support from a student at the University of Sarajevo. This fieldwork was made possible through the combination of funding from the French Embassy in Sarajevo and a research project on geographical indications (the GingKo project3) in partnership with CIRAD4. The main contribution of this work was to provide a more detailed understanding of livestock systems and to initiate reflection on the environmental impacts of the different production systems (Artus & Fraisse, 2024).

44This phase should not be interpreted as a deviation from the initial research intention but rather as a sectoral deepening, focused primarily on agricultural and livestock dynamics. It ultimately allowed us to advance the territorial problematization by adopting new and complementary analytical angles.

An Ongoing Process – Consolidating a Learning Territory

45The various data collection techniques implemented across these phases have served multiple purposes but together have made it possible to assemble a corpus necessary for anchoring environmental intentionality within an increasingly precise understanding of the territory—namely, the karst poljes of Canton 10. Some data are classically used to produce research results, while others are crucial to extending the diagnostic process beyond the lifespan of specific projects or funding cycles. For instance, by understanding why a given partnership is effective, or by identifying gaps in knowledge and the actors interested in addressing them, it becomes possible to envision meaningful follow-up actions. Finally, other elements collected during these missions help determine how and through which channels results should be disseminated so that local actors can quickly appropriate and act upon them.

46These successive phases have progressively defined a learning territory and shaped the problematization of our inquiry. Two main lines of work are now being pursued: (i) exploring the relationships between livestock systems, territorial dynamics, and the prioritization of environmental issues, and (ii) defining co-constructed environmental project initiatives, drawing on the mapping of actor networks and the ongoing discussions around legal statuses and environmental labels.

47In this article, we detail our approach and methodology (Figure 3) to place this territorial and environmental diagnostic in a broader discussion. It is important to emphasize, however, that this remains a work in progress, continuing into 2025 with a regional program on geographical indications and stakeholder engagement activities focused on climate change awareness, linked to the territorial diagnostic and funded by the French Ministry for Europe and Foreign Affairs as well as the European Climate Foundation.

Figure 3: Timeline of our approach and methodological bricolage. The loops emerging from the bottom illustrate the inductive, comprehensive, and iterative logic of the process (according to the authors)

Figure 3: Timeline of our approach and methodological bricolage. The loops emerging from the bottom illustrate the inductive, comprehensive, and iterative logic of the process (according to the authors)

A Territory of Poljes with Specific Hydrological and Agro-Ecological Dynamics

48This section presents the main results of the inquiry, highlighting the most significant elements of the territorial diagnostic, ranging from the delimitation of the study area to the identification of key socio-environmental dynamics—some of which are often overlooked or rendered invisible by environmental NGOs.

49The study area is defined first and foremost by the major karst poljes: Duvanjsko, Glamočko, Kupreško, and Livanjsko, all part of the same watershed located in Canton 10 of Bosnia and Herzegovina, near the border with Croatia. This canton is one of ten in the Federation of Bosnia and Herzegovina, itself one of the two political entities constituting the Bosnian state, alongside the Republika Srpska. Croatian ethno-nationalists have attempted to designate this territory as “Herzeg-Bosnia Canton,” a naming rejected by the Constitutional Court of the Federation on the grounds that it violated the canton’s ethnic diversity (McCormick, 2014). These four poljes, located at different altitudes, are distributed across five municipalities: Bosansko Grahovo, Glamoč, Kupres, Livno, and Tomislavgrad (Figure 4).

Figure 4: The poljes of Canton 10: a Bosnian karst territory within the Dinaric Alps

Figure 4: The poljes of Canton 10: a Bosnian karst territory within the Dinaric Alps

A Karst Territory and a Remarkable Hydro-Geomorphological System

50The hydrology of the poljes is highly distinctive, shaped by the development of subterranean drainage networks associated with karstic erosion. The poljes are connected by underground flows fed by ponors (sinkholes) and estavelles, into which the meandering rivers of the poljes disappear, re-emerging spectacularly as resurgences, often at the base of dramatic limestone escarpments (Figures 5 and 6). This territory is particularly remarkable for its size: Livanjsko polje is reputed to be both the largest periodically flooded polje in the world and the largest in the Dinaric Alps. It covers more than 458 km², stretching 30 km in length and 6 km in width. It has been the subject of study for over a century, notably by Jovan Cvijić, one of the founders of karst geography in the early twentieth century (Cvijić, 1918; Ginsburger, 2015).

51Livanjsko polje is only part of a much larger karst system that spans almost the entirety of the five municipalities. The surface area of this hydro-geomorphological complex makes it unique, encompassing virtually all the characteristic landforms of karst: caves, galleries, shafts, dolines, poljes, ponors, estavelles, resurgences, and rocky escarpments (Figure 6).

52In general, water, whether by surface runoff from the polje slopes, throughflow, or resurgence, collects in the lowest areas of the poljes, forming wetlands or marshes, or disappears into ponors (Figure 5). Owing to their extreme flatness, the poljes host slow-moving, meandering rivers and seasonal flooding (Bonacci, 2013). Another hydrological particularity of this territory is that it straddles two distinct drainage basins: the Cetina basin, which drains toward the Adriatic Sea, and the Sana basin, which ultimately flows into the Sava and the Danube, reaching the Black Sea.

53Historically, local inhabitants dug canals and drains within the poljes and constructed small reservoirs on the plateaus or near springs. The most significant hydraulic interventions, however, occurred during the Yugoslav socialist period, aimed at hydroelectric production (Nicod, 1978). The Buško marsh, located south of Livanjsko polje, was transformed into a lake by sealing ponors, diverting rivers, and constructing a southern embankment (Figures 5 and 6). These infrastructures require maintenance at least every two years, involving the draining of the lake to repair leaks in the limestone bedrock. A subterranean river was also reinforced to feed the Ruda hydroelectric plant (municipality of Otok) in Croatia. Originally conceived within the framework of a unified Yugoslav state, this hydraulic system now spans two independent states (Bosnia and Herzegovina and Croatia).

54This hydro-geomorphological system is thus both unique and remarkable. As in many karst areas (Pisano et al., 2022), the issues of continuity and management of hydrographic networks are central to ensuring water supply, both in terms of quality and quantity, throughout the year. In karst environments, water infiltrates rapidly and carries numerous particles (including pollutants), and it is typically abundant in winter and spring but extremely scarce at the end of summer (Brinkmann & Parise, 2012; De Waele, 2009).

Figure 5: A complex hydraulic system composed of subterranean networks and surface flows

Figure 5: A complex hydraulic system composed of subterranean networks and surface flows

This figure overlays (1) the Yugoslav map of karst water connections in the Cetina basin prepared in Zagreb in 1970 by A. Magdalenić and (2) contemporary georeferenced data on surface water bodies (lakes, wetlands, rivers, canals).

(Legend: utvrđene podzemne veze krških voda u slivu cetine = confirmed underground connections of karst waters in the Cetina basin; ponor = sinkhole; estavela = estavelle; stalni izvor = permanent spring; povremeni izvor = intermittent spring; utvrđena jaka podzemna veza = confirmed major underground connection; utvrđena podzemna veza = confirmed underground connection; problematična podzemna veza = unconfirmed underground connection; granica sliva Cetine = Cetina basin boundary).

Figure 6: Selected landscapes and key features of this karst territory and its hydro-geomorphological system

Figure 6: Selected landscapes and key features of this karst territory and its hydro-geomorphological system

F. Lerin’s photographs.

Land Use

55Each polje contains at least one permanent or seasonal wetland (Figure 4). These wetlands are recognized as habitats that concentrate endemic or rare biodiversity, where naturalists observe remarkable adaptations to seasonal fluctuations in water levels (e.g., fish species that burrow into the soil and aestivate during the dry summer months). They also serve as stopover sites for migratory bird populations, as these poljes are located along major migration corridors linking Northern Europe, the Russian tundra, and sub-Saharan Africa.

56Aside from these wetlands, land use within the poljes is dominated by agricultural areas, which are often underutilized. Using the typology developed by the European Environment Agency and the Corine Land Cover (CLC) database (Corine Land Cover, 2018; Büttner et al., 2021), three main land cover types are found within the poljes: (i) Complex cultivation patterns, characterized by a patchwork of small parcels of diversified annual crops, meadows, or complex permanent crops; (ii) Non-irrigated arable land, including cereals, legumes, fodder crops, row crops, and fallow plots; (iii) Pastures and grasslands, consisting of dense grassy areas primarily grazed, from which fodder may also be mechanically harvested.

57Surrounding the poljes, the landscape is dominated by natural grasslands and three major forested zones located mostly on the surrounding relief (Figure 7).

Figure 7: Land cover and land use in Canton 10

Figure 7: Land cover and land use in Canton 10

European Environment Agency, Corine Land Cover (CLC), 2018.

58Nearly all forested areas and natural grasslands are publicly owned and managed by the Šumarija, decentralized forestry services with offices in each municipality. Forests managed by the Šumarija in Canton 10 are certified under the Forest Stewardship Council (FSC) scheme, an international label guaranteeing sustainable forest management based on ecological, social, and economic indicators. Originating from the Yugoslav forest management system, these forests are uneven-aged and mixed-species and continue to be managed according to what would today be considered sustainable forestry practices—although there are some isolated examples of privately owned, even-aged monoculture plantations.

59Agricultural land is either privately owned or publicly owned and managed by the municipalities (the latter being land that was nationalized for state farms during the socialist period). Due to rural exodus, emigration, and the lack of return migration among younger generations since the war, many private agricultural lands are abandoned, overgrown, or left as open-access commons for the remaining livestock breeders in nearby villages.

Landscape Units

60This combined description of the territory—particularly its hydro-geomorphological system and land cover—enabled us to delineate several landscape units based on vegetation cover and land use: poljes, mountain zones, karst plateaus, and hill zones (Figure 8). In order to identify and describe these landscape units, it was necessary to first delimit the study area. During the second phase of the methodology, the study area was delineated according to municipal boundaries. Municipalities represent a key administrative level in the Federation of Bosnia and Herzegovina, especially regarding agriculture, forestry, hunting, and water management.

61During the third phase, the agrarian diagnostic focused on the geographical indication (GI) of Livanjski Sir cheese considered almost the entirety of Canton 10, without Dvar in the north. According to the GI specification, the production area includes the entire canton, encompassing the five municipalities previously mentioned (Bosansko Grahovo, Glamoč, Kupres, Livno, and Tomislavgrad). This second delimitation also brings together diverse biogeographical and social units.

Figure 8: Distribution of the four main landscape units across the poljes of Canton 10

Figure 8: Distribution of the four main landscape units across the poljes of Canton 10

62This definition of landscape units does not preclude internal variation and sub-units within each major category. For instance, the polje landscape unit can be subdivided into two components: the polje floor, which contains permanent or seasonal wetlands and sometimes canalized watercourses, and the slopes and foothills, which host villages, roads, and rolling hills separating flat areas.

63The mountainous zones include commercially exploitable forests, wooded slopes, high-altitude pastures, saltus (uncultivated land), and rocky areas. The karst plateaus are highly diverse: some are flat, resembling causses and characterized by numerous dolines; others are gently sloping with near-steppe vegetation. Finally, the hill zones may contain wooded hills or canyons of varying openness.

64Describing this territory in terms of landscape units allows the poljes to be situated within a broader and more complex territorial framework, avoiding the risk of treating them in isolation and instead associating them with the surrounding features that separate and border them.

Territorial and Landscape Dynamics

65The inductive, territory- and actor-based approach employed here reveals three territorial dynamics that are largely absent from environmental projects currently proposed for the poljes: (i) large-scale pastoral livestock systems that use the karst plateaus and hill zones, (ii) highly contrasting settlement patterns ranging from ruined or partially abandoned villages to small urban centres with numerous second homes, and (iii) the development of so-called “green” energy infrastructure.

66The territorial diagnostic confirmed the presence of both dairy cattle and sheep systems oriented toward milk production for cheese—part of which is exported to the European Union—but, more unexpectedly, also revealed the existence of large-scale pastoral cattle operations dedicated to beef production. While cheese production and dairy farming are deeply rooted in the history of the territory (dating back at least to the Austro-Hungarian period in the late 19th century), beef cattle operations represent a relatively new and distinctive trend (emerging in the early 2010s) that is not often observed in the Western Balkans. Consequently, understanding the environmental impacts of this type of livestock production on the karst system becomes a key issue for prioritizing environmental concerns and their relationship to economic dynamics.

67It is therefore essential to assess and measure water use and potential contamination, manure management practices, and grazing strategies, while also accounting for the positive effects of pastoral practices on maintaining open landscapes and associated semi-natural biodiversity. Moreover, in the context of climate change, which is increasing the likelihood and intensity of wildfires, pastoralism provides an important tool for combating land abandonment and spontaneous afforestation—both of which are factors that heighten fire risk.

68For this reason, following the initial diagnostic, a more detailed examination of livestock systems was undertaken through the agrarian diagnostic (the third phase of the methodology). Preliminary results of this work indicate the coexistence of two types of systems: intensive dairy production units, likely generating significant negative externalities (depending on housing periods and effluent management), and an increasing number of extensive beef cattle operations relying on large grazing areas.

69These agricultural dynamics are closely tied to the territory’s demographic evolution, which is visible in its landscapes. The extreme diversity of settlements highlights a dual phenomenon: on one hand, depopulation leading to the abandonment of some villages, and on the other hand, the seasonal re-investment of the territory by emigrant populations returning in summer with increased financial resources. Although the war in Bosnia and Herzegovina officially ended in 1995, the scars of conflict remain visible, and minefields have not yet been fully cleared. Entire villages remain abandoned and in ruins, while large new houses are being built in neighbouring settlements.

70These demographic dynamics have multiple causes. First, the region’s recent history, having been traversed by front lines—particularly between Croatian and Serbian troops—has left lasting effects. For example, in the northern part of the Livno municipality, the villages of Provo, Sajković, and Bojmunte were battlegrounds in November–December 1994. Today, these villages remain in ruins, and their population declined significantly between the last two censuses (1991 and 2013) (Figure 9).

71Second, the post-war period, shaped by the Dayton Accords and the complex system of governance (municipalities, cantons, entities, and state) controlled by ethno-political majorities, resulted in ethnic-territorial consolidation (we use the term “ethnic” with caution, as it supplanted the notion of “nation” used during the Yugoslav period; today, “ethno-nationalism” is a common term to describe the political scene). Territories where one of the three main ethno-national groups (Bosniak/Muslim, Croat, Serb—according to national census categories)5 was dominant tended to consolidate populations identifying—or being identified—as belonging to that group.

72For instance, in 1991 the town of Livno had a Muslim majority, within a municipality and canton that were Croat-majority; by 2013, the canton, municipality, and town were all Croat-majority. Many other localities that had been mixed or Serb/Muslim-majority in 1991 are now predominantly Croat. Another example is Kupres municipality, which had a Serb majority in 1991 but a clear Croat majority by 2013 (Figure 9).

73Finally, the rural exodus and significant emigration affecting the entire country over the past three decades have led to sharp population declines between 1991 and 2013 in rural municipalities such as Glamoč and Bosansko Grahovo, with only slight population increases in more urban centres such as Livno, Tomislavgrad, or Kupres (Figure 9). In some villages—Žabljak being a prime example—a new dynamic is visible, with recent housing construction financed by returnees from the European Union (Croatia, Germany, Austria, France, etc.) or from Sarajevo and Mostar. This is a classic phenomenon: these “economic migrants” send remittances to their home territory and to family members remaining in situ (especially the elderly), and they return during the summer and holiday periods.

Figure 9: Population change between 1991 and 2013 (the two most recent censuses) in the municipalities of the learning territory and in selected villages

Figure 9: Population change between 1991 and 2013 (the two most recent censuses) in the municipalities of the learning territory and in selected villages

Bosnian statistical agency. Under the Yugoslav communist regime, up until the 1991 census, the population could self-identify as Muslim (designating a nationality rather than a religion), Croat, Serb, Yugoslav, or other. The latest census in Bosnia and Herzegovina (2013) recognizes four categories: Bosniak (replacing “Muslim”), Croat, Serb, and Other.

74The third territorial and landscape dynamic we identified is the accumulation of infrastructure for so-called “green” or “sustainable” energy production. These infrastructures are the subject of significant controversy, particularly from environmental associations such as CZZS. The territory hosts three types of energy production: wood (and pellet production), water (hydropower), and wind (through wind farms).

75To evaluate the criticism of environmental associations, it is important to note that Bosnia and Herzegovina’s energy mix remains dominated by coal and oil (IEA, 2023). Although the three small coal mines in Canton 10 have been closed, coal—and especially lignite—continues to be extracted elsewhere in the country, and there is currently no planned phase-out of this highly polluting energy source, which is both a major source of greenhouse gas emissions and a crucial fiscal resource for local governments (Beyond Fossil Fuels, 2024), but also detrimental to the health of local population. Bosnia and Herzegovina exports this electricity based on fossil ressources, particularly to Croatia and thus to the European Union (Electricity Map, 2023).

76On this territory, the resources exploited for energy production are primarily destined for export: wood pellets are often transported to Croatia, water from Lake Buško feeds a hydroelectric plant in Croatia, and numerous wind projects are not linked to any plan to phase out coal-fired power plants in the north of the country (Bricault et al., 2023). Furthermore, there does not appear to be a coherent national strategy for the development of these projects, nor a consideration of the specific interests and issues of each territory—sometimes even resulting in contradictory initiatives. For example, at the national level, a project may aim to build small hydroelectric plants on many rivers, while at the local level another project seeks to dismantle dams and certain canals to restore river continuity and aquatic biodiversity (Bricault et al., 2023).

77Environmental associations are also highly critical of the numerous wind farm projects in this area, citing both the scale of the projects and naturalist data identifying the poljes as internationally significant “Important Bird Areas,” particularly for migratory species. Finally, although forestry exploitation and pellet production are carried out by local companies, local actors frequently denounce the fact that wind power projects and hydroelectric production generate very limited financial or economic benefits for the territory.

78Compensation paid by the Croatian company operating the Ruda hydroelectric plant to the municipalities of Tomislavgrad and Livno is widely considered insufficient. As one electricity agency professional in Livno noted:

“The Croatian electricity company, Orlovac, pays each year something like 1 million marks to Livno and 2 million marks to Tomislavgrad. I am not sure—it depends on electricity production. They produce this energy in Croatia, but the lake is in Bosnia and Herzegovina.”

79Similarly, the wind farm under construction in 2023 on the ridgeline separating the Tomislavgrad and Livno poljes is being built by a Chinese company employing foreign labour, thus having virtually no effect on the local labour market, where unemployment remains very high.

The Territory as a Framework for Integrated Environmental Diagnosis?

80Following the presentation of the initial conclusions of the territorial diagnosis, it is now essential to draw some lessons from this process. We shall proceed in two stages: first, by examining the shift in our understanding of the territory, its delineation, and its heritage, development, and environmental challenges; and second, by reflecting on our so-called “embedded” posture, which allows us to deepen our role as knowledge brokers and to initiate, together with stakeholders, the construction of subsequent stages conceived as a process of reciprocal learning derived from the territory and its actors.

Structuring a Learning Territory

81As stated at the outset, our premise is that environmental attributes are the outcome of the territory’s history and must be understood as entities embedded within a dynamic agri-environmental system. Accordingly, the conditions for their continuity and reproduction must be considered, which entails maintaining certain social and economic functions. In order to assess the degree to which these attributes are under threat, it is necessary to analyse the dynamics of the territory and to distinguish between the actors who impact, and those who manage, each attribute.

82By combining ASGE, landscape analysis, and agrarian diagnosis—through a triangulation of semi-structured interviews, bibliographic analyses (scientific articles, statistics, technical reports), in situ observations, and cartographic work—we identified the environmental amenities of the territory and discussed them considering local practices and dynamics. These findings were debated with environmental NGOs, local authorities (municipalities and Canton 10), and economic actors during restitution sessions held in 2023 and 2024.

83Based on this diagnosis, we propose two key shifts compared to the environmental projects that initially brought us into this field: (i) the delineation of the territory and (ii) the centrality of water as a critical issue. One of the guiding principles of this approach was the necessity of not taking the territory as a given, but rather of defining its contours and boundaries in a reasoned and informed manner. Initially, our colleagues from CZZS and their Biodiversity Program had identified the Livno polje (Livanjsko polje) as the core study area, later extending it to the three other major poljes in the region. In their approach, the polje—strictly defined as a closed plain—was taken as the basic unit of the environmental project, within which ornithological, botanical, and herpetological inventories were carried out.

84Building on these contributions, we considered the four poljes as interconnected units forming a larger ensemble that necessarily included intermediate spaces. This led us to identify and describe hill areas, slopes and mountain ranges, saltus zones, herbaceous and forest formations, smaller poljes, as well as karstic plateaus riddled with dolines and scattered with lapiaz. We concluded that it was this broader biogeographically coherent entity, rather than individual poljes, that should be considered the relevant scale of analysis. Our approach thus diverges from the ongoing initiative to designate Livanjsko polje alone as a national protected landscape (Cener21, 2021; Jaćimovska, 2019), an endeavour stalled by local political disputes and municipal blockages (Bricault et al., 2023).

85Furthermore, this inductive approach—supported by extensive engagement with territorial management actors—highlighted the primacy of water as the determinant issue at the top of the hierarchy of environmental challenges. We drew upon hydrological studies conducted during the socialist Yugoslav period, which remain the only systematic descriptions of these watersheds and of the surface and subterranean flows characteristic of karst landscapes. Issues of water quality and hydrological continuity—remarkable for the scale of underground structures and the diversity of surface features (ponors, springs, estavelles, wetlands, etc.)—emerge as central. This challenge, typical of karst environments, aggregates multiple themes defended by environmental NGOs, including the protection of migratory birds in wetlands, endemic biodiversity in seasonally flooded areas, energy concerns related to small dams and water diversions, and aquatic biodiversity. It also enables a comprehensive territorial approach encompassing not only the four major poljes but also the surrounding and intermediate landscapes.

86The territorial diagnosis, grounded in the combination of multiple analytical frameworks, thus enabled us both to transcend disciplinary silos to define a potential environmental project territory, and to reveal gaps in data and knowledge concerning specific attributes and local dynamics. It became apparent that agricultural dynamics—particularly livestock farming—were largely underestimated, if not ignored, by the locally engaged environmental NGOs, often portrayed reductively as having solely negative impacts. A detailed analysis of agrarian systems was therefore necessary, considering not only the poljes themselves but also the use of intermediate spaces, karst plateaus, hill areas, slopes, and mountain pastures. This analysis allowed us to assess both the negative and positive externalities of each production system and to identify levers for supporting certain pastoral and extensive systems (through subsidies, or improved access to public lands).

87This underestimation of the agricultural sector in shaping landscape and environmental dynamics undermines the co-construction of a territorial strategy, particularly given its economic vitality—most notably in dairy production, but increasingly in cheese and meat production as well. No environmental territorial strategy can be constructed without engaging these producers, just as no local authority can credibly embark on a territorial project without their involvement, given that they, alongside the forestry sector, constitute the principal economic activity of the canton.

A Territorial Approach and Knowledge Brokerage

88The embedded and brokerage posture developed in this territorial diagnosis is neither prescriptive nor instrumental. Rather, it is situated within a logic of supporting transitions, wherein the diagnosis serves as a lever to render visible opportunities, tensions, margins of manoeuvre, and potential axes of collective action. Our embeddedness thus evolves by and for the territory, structured around three dimensions: (i) the environmental interests of the territory (water, landscape, etc.); (ii) the actors to be enrolled (farmers, canton); and (iii) the institutional frameworks and scales of action to be mobilized (territorial project, landscape mosaic rather than polje).

89The first dimension of brokerage consists of drawing upon existing knowledge stocks and data gathered through fieldwork and having them discussed and validated by the actors to be enrolled, who are also potential co-constructors of subsequent stages. Although this positioning is not always readily understood, the practice of inquiry, the time invested, the restitution sessions, and the partial integration of diverse positions and interests facilitate the acceptance of such an attitude. This has allowed progress, for example, on the apparently formal issue of territorial delineation (for action), consolidating with local actors the rationale underpinning this delineation.

90Accordingly, the territorial approach makes it possible to establish a mapping of actors and to adapt brokerage strategies to the existing configurations of power and influence. In some cases, this entails relying on actors in leadership positions; in others, on those most impacted by environmental changes; or, alternatively, developing a prospective outlook to engage emerging actors considering local socio-economic and political dynamics. The example of pastoral farmers, discussed above, is illustrative, as they are generally convinced that their practices shape the reputation of their products, attributing their distinctiveness and quality to the preserved environment.

91As knowledge brokers, our role also consists in bridging worlds that seldom communicate farmers and policymakers, local administrations and donors, NGOs, and so forth. Restitution sessions thus serve as arenas for dialogue among territorial actors who otherwise have few opportunities to interact. Linking with European and international norms and tools is another dimension of this brokerage work, involving scale-shifts and multi-actor inquiries—for instance, through European pre-accession instruments, geographical indications as market tools, or heritage designations. The implicit premise of this approach is that while some precision may be lost, relevance is gained—provided that specific knowledge is rigorously integrated and that the strategic intent is co-constructed and made explicit.

92The effects of such a diagnosis, grounded in an inductive methodology, on the shaping of a learning territory can be observed on two levels. First, the empowerment of local actors: by involving them from the stage of issue definition (e.g., farmers, NGOs, local administrations in territorial delineation and issue prioritization), or by legitimizing their empirical and historical knowledge of the territory (e.g., recognizing the role of herders in managing and maintaining landscapes). Second, consequences for environmental actors: some revisit their priorities (e.g., CZZS and EuroNatur now considering greater integration of farmers into their projects); others envision new alliances (such as between naturalist NGOs and local administrations); and still others emerge as significant actors (in the polje region, notably Canton 10 and its agricultural and environmental Ministries, which have seized upon this diagnosis as an opportunity to build territorial environmental policies).

93Nevertheless, the dissemination and appropriation of this diagnosis face three main limitations. First, key actors such as municipalities remain reluctant to move beyond the scale of individual poljes. Second, given the inductive and interpretive nature of the approach, it is inherently difficult to predetermine expected outcomes and limitations. Some actors unaccustomed to such approaches may thus be disappointed when research and diagnosis do not immediately translate into concrete projects or recommendations. Finally, certain significant issues raised by the diagnosis are not appropriated by local actors due to entrenched historical and political blockages—for instance, ethnic issues, which, despite individual attempts at transcendence, remain overshadowed by broader administrative, political, and memorial constraints.

Conclusion

94We conclude by focusing more specifically on the question of potential territorial and environmental labels that could support a strategy of dynamic conservation for this unique territory. As discussed, hydrology emerges as a fundamental dimension of these karst landscapes. The stratified system of poljes and the mountain chains separating them form a remarkable—arguably unique—hydrographic complex within the Dinaric Alps, thereby creating notable conditions for future labelling initiatives that could reinforce the protection of its functioning as an integrated system encompassing both surface and subterranean waters. A UNESCO Global Geoparks label could be considered6.

95A second conclusion from the territorial diagnosis is that the region’s historical trajectory and current situation qualify it as a cultural landscape of Mediterranean pastoralism. In this respect, pursuing a potential UNESCO designation as a “living, evolving cultural landscape” appears promising (like the UNESCO designation of Causses and Cévennes, a cultural landscape of Mediterranean agropastoralism in another karstic hinterland7). Discussions on such a possibility have already been initiated with local actors and the Canton authorities, who have expressed strong interest. Moreover, the preparation of such a dossier would necessitate a co-construction agenda involving both national and international scientific expertise and local stakeholders, thereby providing the backbone of a territorial project.

96Our initial posture—defining the territory and establishing a hierarchy of environmental issues—thus finds its justification and relevance in these conclusions and propositions, which reflect both the universal and exceptional character of the identified heritage and the local economic dynamics and actor constellations, potential stakeholders in a territorial project. It was crucial not to approach the territory through a narrow sectoral lens, which typically constrains thought within silos defined by keywords (adaptation, migratory corridors, greenhouse gases, wetlands, rural development, minorities, gender, etc.). While sectoral visions have their merits for specific projects and for fostering collaboration among environmental actors, they risk constraining action within a theme and preventing the development of an integrated vision—that is, a territorially grounded approach informed by diverse field methods and data.

97The territorial diagnosis and knowledge brokerage approach we propose requires methodological flexibility and dynamism in the research posture. These elements, often difficult to reconcile with current research procedures (project-based funding, detailed ex ante definition of objectives, steps, and expected outcomes), oblige us to resort to “bricolage” to sustain our inquiries, to maintain a long-term presence in the field, and to accept conceptual and methodological dynamism, instability, and flexibility.

Haut de page

Bibliographie

Akrich, M., Callon, M., & Latour, B. (2006). Sociologie de la traduction : Textes fondateurs. Paris : Presses des Mines.

Artus, E., & Fraisse, A. (2024). Agrarian system and landscape dynamics in the production area of Livanjski sir, Bosnia and Herzegovina. Montpellier : Institut Agro.

Ballouche, A. (2016). Forêts versus savanes : l’héritage d’un forestier colonial en écologie et géographie tropicale de l’Afrique de l’Ouest. Dans D. Chartier & E. Rodary (dir.), Manifeste pour une géographie environnementale : Géographie, écologie, politique (pp. 184-204). Paris : Presses de Sciences Po.

Bernard, C., Poux, X., Herzon, I., Moran, J., Pinto-Correia, T., Dumitras, D. E., Ferraz-De-Oliveira, M., Gouriveau, F., Goussios, D., Jitea, M. I., Kazakova, Y., Koivuranta, R., Lerin, F., Ljung, M., Lomba, A., Mihai, V. C., Puig De Morales Fusté, M., & Vlahos, G. (2023). Innovation brokers in High Nature Value farming areas: A strategic approach to engage effective socioeconomic and agroecological dynamics. Ecology and Society, 28(1), 20. https://doi.org/10.5751/ES-13522-280120

Bernard-Mongin, C. (2016). La biodiversité par projet : Réflexivité engagée et dispositif stratégique en Albanie (thèse de doctorat, Université de Montpellier).

Bertrand, C., & Bertrand, G. (2002). Une géographie traversière : L’environnement à travers territoires et temporalités. Versailles : Éditions Quae.

Beyond Fossil Fuels. (2024). National coal phase-out commitments in Europe: Coal phase-out map. https://beyondfossilfuels.org/coal/

Blanc-Pamard, C. (1986). Dialoguer avec le paysage ou comment l’espace écologique est vu et pratiqué par les communautés rurales des hautes terres malgaches. Dans Y. Chatelin & G. Riou (dir.), Milieux et paysages (pp. 17-36). Paris : Masson.

Bonacci, O. (2013). Poljes, ponors and their catchments. Dans J. F. Shroder & A. Frumkin (dir.), Treatise on geomorphology (Vol. 6, Karst geomorphology, pp. 112-120). San Diego : Academic Press.

Bricault, P., Daras Siffait De Moncourt, B., Debono, S., Durand, T., Fiévet, A., Fresel, L., Naudin, A., Kouaho, H., Legay, J., Montillot, G., Piechon, C., Piedigrossi, L., Pouthé, G., Pretesacque, J., & Sanchez, B. (2023). Environmental-territorial diagnosis of polje’s watershed systems in Bosnia and Herzegovina (BiH): Case study of four karst poljes in Canton 10 (Duvanjsko, Glamočko, Kupreško and Livanjsko). Montpellier : AgroParisTech – Association AIDA. https://www.courrierdelaplanete.org/wp-content/uploads/Rapp_mission_coll_2023_VF.pdf

Brinkmann, R., & Parise, M. (2012). Karst environments: Problems, management, human impacts, and sustainability. An introduction to the special issue. Journal of Cave and Karst Studies, 74, 135-136.

Brunet, R., Ferras, R., & Théry, H. (2005). Les mots de la géographie : Dictionnaire critique. Montpellier-Paris : Reclus – La Documentation française.

Büttner, G., Kosztra, B., Maucha, G., Pataki, R., Kleeschulte, S., Hazeu, G., Vittek, M., Schröder, C., & Littkopf, A. (2021). Copernicus land monitoring service CORINE land cover – User manual. European Environmental Agency, Copernicus Land Monitoring Service. https://land.copernicus.eu/user-corner/technical-library/clc-product-user-manual

Callon, M. (1986). La domestication des coquilles Saint-Jacques et des marins-pêcheurs dans la baie de Saint-Brieuc. L’Année sociologique, 36, 169-207.

Callon, M., Lascoumes, P., & Barthe, Y. (2001). Agir dans un monde incertain : Essai sur la démocratie technique. Paris : Le Seuil.

Cener21. (2021). Scientific justification for the proclamation of the protected area definition category 5 protected landscape (rapport).

Cholley, A. (1946). Problèmes de structure agraire et d’économie rurale. Annales de géographie, 298, 81-101.

Cochet, H. (2011a). L’agriculture comparée. Versailles : Éditions Quae.

Cochet, H. (2011b). Origine et actualité du « système agraire » : Retour sur un concept. Revue Tiers Monde, 207(3), 97-114. https://doi.org/10.3917/rtm.207.0097

CORINE Land Cover (vector). (2018). Europe, 6-yearly – Version 2020_20u1 (May 2020). https://doi.org/10.2909/71c95a07-e296-44fc-b22b-415f42acfdf0

Cottet, M. (2019). Notion en débat : Paysage. Géoconfluences. https://geoconfluences.ens-lyon.fr/informations-scientifiques/a-la-une/notion-a-la-une/paysage

Crozier, M., & Friedberg, E. (1977). L’acteur et le système. Paris : Éditions du Seuil.

Cvijić, J. (1918/2014). La péninsule balkanique : Géographie humaine. Belgrade : Bibliothèque nationale de Serbie / Paris : BnF.

Davodeau, H. (2005). La sensibilité paysagère à l’épreuve de la gestion territoriale. Cahiers de géographie du Québec, 49(137), 177-189. https://doi.org/10.7202/012299ar

Décamps, H., & Décamps, O. (2007). Organisation de l’espace et processus écologiques. Économie rurale, 297-298, 55-70. https://doi.org/10.4000/economierurale.1990

Demailly, K.-E. (2017). Book review: Manifeste pour une géographie environnementale. Geographica Helvetica, 72, 1-3. https://doi.org/10.5194/gh-72-1-2017

Electricity Map. (2023). Electricity Maps ApS – Zone : Europe, Bosnia-Herzegovina. Copenhagen. https://app.electricitymaps.com/zone/BA

Feildel, B. (2016). Le bricolage et l’herméneutique comme principes méthodologiques dans l’enquête en sciences sociales : Un outil cartographique pour révéler la relation affective aux espaces de vie. ESO Travaux et documents, 41, 89-99. https://shs.hal.science/halshs-01409619

Funtowicz, S. O., Ravetz, J. R., & Jerome, R. (1993). Science for the post-normal age. Futures, 25(7), 739-755.

Gasnier, M., Guinchard, C., Kroichvili, N., Masselot, C., Moine, A., Nuninger, L., & Winckel, N. (2022). Jalons pour une construction collaborative du concept de revitalisation territoriale. Cybergeo: European Journal of Geography, 1007. https://doi.org/10.4000/cybergeo.38404

Gibbons, M., Limoges, C., Nowotny, H., Schwartzman, S., Scott, P., & Trow, M. (1994). The new production of knowledge: The dynamics of science and research in contemporary societies. London : Sage Publications.

Ginsburger, N. (2015). Les Balkans avec ou sans Cvijić : Géographes et géologues universitaires austro-allemands, français et serbes dans un espace européen périphérique. Dans P. Clerc & M.-C. Robic (dir.), Des géographes hors-les-murs ? Itinéraires dans un monde en mouvement (1900-1940) (pp. 323-354). Paris : L’Harmattan.

Guinchard, C., Kouassi, G.-N., & Moine, A. (2023). Jalons pour une géographie de la traduction : Mieux comprendre le territoire du lac de Buyo. VertigO, 23(2), 1-32.

IEA. (2023). Energy statistics data browser. Paris : IEA. https://www.iea.org/data-and-statistics/data-tools/energy-statistics-data-browser

Jaćimovska, M. (2019). Spatial protection of Livanjsko polje in the framework of the UNEP/GEF project: Achieving biodiversity conservation through creation, effective management and spatial designation of protected areas and capacity building. Dans P. Sackl, S. Ferger, N. Sarajlić, D. Kotrošan, & G. Topić (dir.), Dinaric Karst poljes : Nature conservation and rural development (pp. 9-15).

Joliveau, T., Michelin, Y., & Ballester, P. (2008). Éléments de méthode pour une médiation paysagère dans un contexte rural. Dans T. Brossard & J.-C. Wieber (dir.), Paysage et information géographique (pp. 257-285). Paris : Lavoisier.

Lerin, F. (2015). The scale issue in global international environment governance: for a transdisciplinary perspective. Dans O. Charnoz, V. Diaz Pedregal et A. L. Kolata (dir.), Local Politics, Global Impacts. Steps to a multidisciplinary of scales (pp. 39-63). Ashgate

Les chercheurs ignorants. (2015). Introduction. Dans Les recherches-actions collaboratives : Une révolution de la connaissance (pp. 11-15). Rennes : Presses de l’EHESP. https://doi.org/10.3917/ehesp.lesch.2015.01.0011

Malešević, B., & Siegel, Z. (2023). Comprehensive territorial approach and environmental stakes in two municipalities of Bosnia and Herzegovina : Mrkonjić Grad and Sanski Most. Montpellier : Association AIDA. https://www.courrierdelaplanete.org/wp-content/uploads/Malesevic_Siegel_report_english_compressed.pdf

Marty, P., Lepart, J., & Caplat, P. (2006). Géographie et écologie des paysages : Quelles relations ? Bulletin de l’Association de Géographes Français, 83(3), 355-367. https://www.persee.fr/doc/bagf_0004-5322_2006_num_83_3_2521

Mazoyer, M., & Roudart, L. (1997). Histoire des agricultures du monde : Du néolithique à la crise contemporaine, Editions du Seuil, 545 p.

McCormick, R. (2014). Croatia under Ante Pavelić : America, the Ustaše and Croatian genocide. Londres : I. B. Tauris.

Mermet, L. (2014). Conserver et restaurer la biodiversité : Un problème d’action stratégique. Dans M. Gauthier-Clerc, F. Mesleard, & J. Blondel (dir.), Sciences de la conservation (pp. 223-229). Louvain-la-Neuve : De Boeck Supérieur.

Mermet, L., Billé, R., Leroy, M., Narcy, J., & Poux, X. (2005). L’analyse stratégique de la gestion environnementale : Un cadre théorique pour penser l’efficacité en matière d’environnement. Natures Sciences Sociétés, 13(2), 127-137.

Moine, A. (2006). Le territoire comme un système complexe : Un concept opératoire pour l’aménagement et la géographie. L’Espace géographique, 35(2), 115-132. https://doi.org/10.3917/eg.352.0115

Nassauer, J. I. (1992). The appearance of ecological systems as a matter of policy. Landscape Ecology, 6(4), 239-250.

Nicod, J. (1978). Les eaux et l’aménagement des poljés du karst dinarique. Méditerranée, 32(1-2), 85-104. https://www.persee.fr/doc/medit_0025-8296_1978_num1_1772

Nogueira, L. A., Bjørkan, M., & Dale, B. (2021). Conducting research in a post-normal paradigm: Practical guidance for applying co-production of knowledge. Frontiers in Environmental Science, 9.

Pisano, L., Zumpano, V., Pepe, M., Liso, I. S., & Parise, M. (2022). Assessing karst landscape degradation: A case study in Southern Italy. Land, 11(1842), 1-18. https://doi.org/10.3390/land11101842

Raffestin, C. (2003). Penser et classer dans les sciences humaines. Revue européenne des sciences sociales, 41(127), 71-85. https://doi.org/10.4000/ress.508

De Waele, J. (2009). Evaluating disturbance on Mediterranean karst areas: The example of Sardinia (Italy). Environmental Geology, 58, 239-255.

Haut de page

Notes

1 For more information: URL: https://czzs.org

2 For more information: URL: www.courrierdelaplanete.org

3 For more information: URL: https://fr-gingko.hub.inrae.fr/

4 Agricultural research Centre for the sustainable development of tropical and Mediterranean regions, for more information: URL: https://www.cirad.fr/

5 For more information: Statistika, URL: http://www.statistika.ba/

6 For more information: URL: https://www.unesco.org/fr/iggp/geoparks/about

7 For more information: URL: https://causses-et-cevennes.fr/

Haut de page

Table des illustrations

Titre Figure 1: Articulation of the three analytical frameworks used in this study (according to the authors)
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-1.png
Fichier image/png, 463k
Titre Figure 2: Schematic representation of the actor system developed during this diagnostic phase (Bricault et al., 2023, p. 63)
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-2.png
Fichier image/png, 546k
Titre Figure 3: Timeline of our approach and methodological bricolage. The loops emerging from the bottom illustrate the inductive, comprehensive, and iterative logic of the process (according to the authors)
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-3.png
Fichier image/png, 347k
Titre Figure 4: The poljes of Canton 10: a Bosnian karst territory within the Dinaric Alps
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-4.jpg
Fichier image/jpeg, 515k
Titre Figure 5: A complex hydraulic system composed of subterranean networks and surface flows
Légende This figure overlays (1) the Yugoslav map of karst water connections in the Cetina basin prepared in Zagreb in 1970 by A. Magdalenić and (2) contemporary georeferenced data on surface water bodies (lakes, wetlands, rivers, canals).
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-5.jpg
Fichier image/jpeg, 571k
Titre Figure 6: Selected landscapes and key features of this karst territory and its hydro-geomorphological system
Crédits F. Lerin’s photographs.
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-6.png
Fichier image/png, 1,8M
Titre Figure 7: Land cover and land use in Canton 10
Crédits European Environment Agency, Corine Land Cover (CLC), 2018.
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-7.jpg
Fichier image/jpeg, 641k
Titre Figure 8: Distribution of the four main landscape units across the poljes of Canton 10
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-8.jpg
Fichier image/jpeg, 634k
Titre Figure 9: Population change between 1991 and 2013 (the two most recent censuses) in the municipalities of the learning territory and in selected villages
Crédits Bosnian statistical agency. Under the Yugoslav communist regime, up until the 1991 census, the population could self-identify as Muslim (designating a nationality rather than a religion), Croat, Serb, Yugoslav, or other. The latest census in Bosnia and Herzegovina (2013) recognizes four categories: Bosniak (replacing “Muslim”), Croat, Serb, and Other.
URL http://journals.openedition.org/vertigo/docannexe/image/50546/img-9.png
Fichier image/png, 300k
Haut de page

Pour citer cet article

Référence électronique

Orianne Crouteix, François Lerin et Jérémy Vendé, « From territorial diagnosis to strategic intent. Knowledge brokerage and co-construction. Fieldwork Insights from Karst Landscapes – the poljes of Canton 10 (Bosnia and Herzegovina) »VertigO [En ligne], 25-2 | Octobre 2025, mis en ligne le 24 octobre 2025, consulté le 19 janvier 2026. URL : http://journals.openedition.org/vertigo/50546 ; DOI : https://doi.org/10.4000/15erj

Haut de page

Auteurs

Orianne Crouteix

PhD in Geography, Member of the TELEMMe Joint Research Unit (UMR 7303), Aix Marseille University, CNRS, France, email address: crouteixo@gmail.com

François Lerin

PhD in Economics, International Association for the Development of the Agri-Environment (AIDA), Environmental and Social Scientist, Combaillaux, France, email address: francois.lerin@posteo.net

Jérémy Vendé

Research Engineer, AgroParisTech, Training and Research Unit G-ENV (Environmental Management), Montpellier, France, email address: jeremy.vende@agroparistech.fr

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search