Introduction

Rangelands in southern Tunisia suffer from severe degradation due to multiple socioeconomic changes, including in pastoral and agropastoral societies, land tenure systems, and mechanization (Nefzaoui and Mourid 2008; Nefzaoui et al. 2012; Ouled Belgacem 2018). These changes have led to increasing conflicts concerning pastoral resources, especially in the context of collective rangelands, owned and managed collectively, thus involving some additional difficulties for their good governance (William et al. 2003). Overgrazing is now the main anthropic factor leading to decline and degradation of the perennial plant cover. The increase in livestock numbers, expansion of cultivated areas at the expense of rangelands, and weakening of traditional grazing systems (transhumance and nomadism), which historically allowed for grazing deferment and control of livestock grazing, have led to the emergence of new sustainability challenges for pastoral resources. Overgrazing is more harmful when coupled with climatic aridity (mean annual rainfall < 150 mm) and may disturb rangeland ecosystems (Ouled Belgacem et al. 2006; Tarhouni et al. 2006).

Ancestral agropastoral societies in Tunisia have developed their local knowledge and strategies for coping with drought and climate variability: mobile or transhumant grazing practices that reduce the risk of having insufficient forage in any location, feed storage during favorable years or seasons, reciprocal grazing arrangements with more distant communities for access to their resources in drought years, and adjustment of flock sizes and stocking rates as the rainy season unfolds, to best match available grazing resources (Hazell and Wood 2007). One of the most well-known and widely practiced traditional techniques for rangeland management in Tunisia and elsewhere is the resting technique (Nefzaoui and Mourid 2008; Squires et al. 2009; Gamoun et al. 2018; Ouled Belgacem 2018). This technique is locally known as “Gdel.” It consists of leaving part of the rangeland to rest (without grazing) for a definite period of 2–4 yr depending on the ecosystem capacity to recover and on climatic conditions, with the aim to reconstitute plant cover, and can result in increases in fodder production, soil organic matter, and biodiversity and reduction in soil erosion (Ouled Belgacem 2012).

However, the socioeconomic and institutional changes to ancient pastoral systems have led to a gradual abandonment of the Gdel technique by pastoral societies. In fact, Tunisian rangelands were managed and controlled before the French occupation (1881) by traditional institutions, called “Myâad,” composed of tribal chiefs, who enjoyed effective power for annual management (opening and closing) of collective rangelands (Nefzaoui and Ben Salem 2011; Gamoun et al. 2018). The Myâad was transformed during the French occupation of Tunisia into a more formal structure called "Land Management Council" (CG), composed mainly of landowners. After independence (1956), this traditional institution was consolidated. Since the 2000s, the government has implemented a complementary and new organization called the Agricultural Developing Grouping (GDA), which comprises a group of landowners and land users. The relationship between the two organizations was marked by time by some conflictual or collaborative actions as appropriate and depending on context (Gamoun et al. 2011).

The progressive decline of rangeland productivity in these contexts of socioeconomic, institutional, and climate changes, as well as the ineffective models of entrusting management of natural resources to government agencies or devolving natural management to user groups and local institutions, are then accompanied with disruption of institutional arrangements for rangeland management, which is reflected by difficulties in cooperation among an increasing number of stakeholders and weak performances of existing farmers' organizations (Robinson et al. 2017; Ouled Belgacem 2018).

Because of their spatial scale and multiple users, owners, and managers, rangelands in Tunisia require some form of coordinated regulation. Collective action is then essential to resolve governance problems since it leads to stronger, more cohesive, and more stable communities (CAPRi 2010). It strengthens local governance institutions and improves rangeland productivity and management (Flintan et al. 2019). However, the success of collective action is highly dependent on actors' behavior and the state of local rangeland governance. These criteria are particularly important under complex and different land tenure systems where many (collective) landowners, land users, and land managers (governance structures) are intervening for decision making on the rangeland resource.

The questions here are under what form and conditions will collective action emerge and be strong enough to manage these rangelands, and what can be done to increase the likelihood of cooperation in the context of devolution programs?

Considerably more attention is needed to the factors that make these stakeholders willing and able to take on an expanded role (capacity for cooperation) in a context of effective collective action, which usually includes rules on using (or refraining from using) a resource, as well as processes for monitoring, sanctioning, and dispute resolution.

Considerable research has been devoted to this topic and gives attention to the "enabling environment" for effective collective action on rangeland management.

Wheatland and Chêne (2015) define collective action as a variety of actions that are undertaken by individuals or groups or an organization that is acting on their behalf to achieve a collective purpose or shared interest of the group or individuals. As an analytical concern, collective action is a concept that is associated with a myriad of disciplines and contexts, which include economic, political, labor, development, agriculture, environment, social, governance, and anticorruption contexts to name but a few. It is a concept that is relevant and adaptable to different contexts and levels (Vanni 2014).

The theory of collective action and its relevance to social issues has been changing over time. Olson's theory of collective action (Olson 1965), the Prisoner's Dilemma and the tragedy of the commons, tended to analyze common resource systems using similar assumptions about actors' behavior and the structure of the decision-making environment (Sally 2000). These assumptions include the homogeneity of actors' assets, skills, discount rates, and cultural views; the possession of complete information about their own benefits and costs; and the decision making based on maximization of material benefits to self over the short term. Game theorists generally recognized refinements focused on the implications of repeated interactions, the possibilities for using contingent strategies that reward cooperation and punish defection to achieve higher levels of cooperation. During the 1970s and 1980s, a framework for studying institutions, related to both natural and man-made resources, called the Institutional Analysis and Development (IAD) framework, was developed (Kiser et al. 1982). This framework considered a metatheoretical conceptual map, which identifies an action arena, patterns of interactions, and outcomes and is structured by attributes that affect the preferences, information, strategies, and actions. It provides a set of analytical tools that link game theoretic and empirical work using multiple methods on collective action (Kiser et al. 1982). From the 1990s, an intensive empirical research and theoretical development raised awareness of empirical and theoretical challenges to the conventional theory, as well as existing conceptual and theoretical refinements, across disciplinary boundaries and provided conceptual and analytical tools that facilitated new research. In particular, the NRC panel increased awareness of the prevalence of common property, of terminology that more clearly distinguishes between the characteristics of the natural resource and the system of property rights, and of the IAD framework for analyzing collective action for the development of institutions such as property rights. The success of Ostrom's research program is especially revealing in this matter. Ostrom uses institutional analysis to examine different ways—both successful and unsuccessful—of governing the commons. In contrast to the proposition of the tragedy of the commons argument, common pool problems are sometimes solved by voluntary organizations rather than a coercive state. Among the cases considered are communal tenure in meadows and forests, irrigation communities and other water rights, and fisheries (Ostrom 2009).

Much of the emphasis here has been on the organizations and regulations taking in account actors' behavior and institutional arrangements. Less attention has been given to the property rights and power force of actors or groups of actors, which are perhaps the most critical factors in enabling the organizations to operate. Property rights play a central role in the management of natural resources. They are often broadly classified as common (held by a community or group of users) and private (held by individuals) (Svendsen 1997).

Shocks and conflicts on unregulated open access common pool resources may divide a community and lead to the overuse and destruction of these resources. This is true if collective action involves only part of the community or dominant social group and results in the exclusion of others from the activities and benefits of the action. Widespread use of the term common pool resources across the social sciences suggested that these resource systems were always governed under a form of property called common property (CAPRi 2010).

This study suggests that power relations must be considered in decision making, in addition to information, skills, and other decision attributes for an effective collective action on rangeland management. In other words, taking into account the limited decision-making power of some stakeholders in coordination mechanisms can strengthen collective action and set the stage for more collaborative interaction between all stakeholders involved in rangeland management. The implementation of this effective collective action on rangeland restoration depends on fully understanding the relationships among institutions and organizations and the ways in which governance powers are distributed in management processes.

This collective action must also consider the differences in land tenure and governance regimes, given the great difference between these regimes in terms of management structures and the way in which resources are managed and allocated.

The implementation of social research to assess these relationships and interactions among governance structures and actors in relation to rangeland restoration could provide more insights concerning sustainable rangeland management in each land tenure system (Giordano et al. 2004; Valkering et al. 2004; Giupponi and Giannini 2010; Ingold 2011, 2013).

Methodological framework

Contextual analysis: Characteristics of the case study

Results

Mapping stakeholders, challenges, and objectives

The starting point of our analysis is to build an initial map of relevant stakeholders involved in the rangeland restoration process (Table 1). The focus group discussions (FGDs) allowed identifying issues and challenges of rangeland restoration based on local experience. The main challenges identified by stakeholders are related to land tenure (e.g., difficulties of clarifying land property acts and accessibility to rangelands), rangeland degradation, drought and water scarcity, and increased cultivated arable area at the expense of highly productive rangelands. Other challenges are related to the high level of installation and maintenance costs of pastoral infrastructure (e.g., water points); social capital of local communities, which constrains collective action; lack of valorization of research findings; lack of monitoring and good implementation of development (investment) projects; and lack of management and technical skills of staff of local administration and organizations. Hence, all outlined factors undermine scaling and limit adoption of the Gdel practice.

Fig. 1.

Typology of rangelands in Tataouine Governorate (Source: own elaboration, adapted from Bel Fkih and Jarray 2018).

Table 1

List of main stakeholders involved in rangeland restoration under different land tenure systems in Tataouine.

Table 2

List of main challenges and objectives of rangeland resting technique (Gdel) under different land tenure systems in Tataouine.

Fig. 2.

Plan of influences and dependencies among main involved stakeholders in collective (on the left) and private (on the right) rangeland restoration process in Tataouine (Source: own elaboration from MACTOR analysis, 2019).

The previous assessment of FGDs allows the identification of five main objectives of land restoration with regard to specific challenges (Table 2). Three objectives are common to both collective and private tenure systems: conservation and restoration of vegetation cover and pastoral systems (O1), sustainability of pastoral resources (O2), and improving the productivity of rangelands (O3). An additional objective of stakeholders in collective rangelands is adaptation of pastoral systems to climate change (O4), while the last objective of actors involved in private land restoration is establishment of a rational exploitation plan of rangelands and their scaling up to larger areas (O5).

Stakeholders' strategic orientations in terms of rangeland restoration

Relationships, mutual influences, and dependencies among stakeholders

The first output from MACTOR analysis reflects the reciprocal influences exerted among stakeholders and their level of dependency on each other for rangeland restoration (Fig. 2). For rested areas in collective rangelands, analysis shows that there are dominant stakeholders who have a strong influence on the others without being strongly influenced (e.g., UTAP and IRA). The analysis also shows that most key stakeholders (GDA, CG, CRDA, and PRODESUD) have a weak influence on the acceptance of the Gdel technique by farmers who are strongly influenced by social norms, culture, and the method of governance of this common pool resource.

The situation differs in the case of private rangelands, where there is a predominance of relay stakeholders who could bridge the gaps among other stakeholders, except for IRA, which is neither influential nor dependent (autonomous). The beneficiaries (owners and users of rangelands) are the most influential (e.g., in terms of decisions related to management and access to rested areas) and most dependent (in terms of access to information and technology). This means that beneficiaries, despite their interest, will be unable to implement restoration without technical and institutional assistance from third-party stakeholders (e.g., IRA and CRDA). This is particularly because of the nature of restoration/resting, which requires specific knowledge on how and when to open rangelands for grazing. In addition, there is a need for regulating agents who can mediate the collective action of all beneficiaries.

Results also reveal the absence of absolute dominant and dominated stakeholders in the private rangeland context. This situation leads to a much more stable rangeland restoration process since the majority of stakeholders have the same power in terms of influence and dependence.

In terms of relation powers (influence in the rangeland restoration decision process), findings show that for collective rangelands, UTAP (farmers' syndical union), IRA (research organization), and public authorities (administration) have an R score higher than the other stakeholders and are then considered as the most powerful (Fig. 3). The GDAs are completely dominated, with a weak R score. In fact, the heavy presence of the administration (CRDA) creates a strong dependency of local associations. In this case, the challenge is the difficulties of acquiring independency and capacities of dialog between CRDA and GDA. The flexibility of the administration is thus important for better performance of rangeland management. The accountability and transparency between stakeholders are also symptoms of effective rangeland governance. The dominance of UTAP is explained by the transformation of the institutional setup of this organization after 2011 (i.e., the Tunisian revolution). This farmers' union has an important role in defending the interests of farmers and breeders and arbitrating their conflicts in order to better manage rangeland restoration.

Fig. 3.

Histograms of power relations of main involved stakeholders in collective (on the left) and private (on the right) rangeland restoration processes in Tataouine (Source: own elaboration from MACTOR analysis, 2019).

In the private rangelands, nearly all stakeholders have the same balance of power. This confirms our previous finding on the stability of the Gdel process in this land tenure system. Despite its mandate for developing livestock feeding strategies, including the provision of technical skills for setting rangeland resting, the OEP is in a nonfavorable balance of power in terms of influence.

Potential alliances and conflicts between stakeholders: Convergence and divergence regarding effective restoration objectives

The MACTOR analysis can also provide potential alliances and conflicts among stakeholders regarding restoration objectives. This can be depicted by identifying the levels of convergence (sum of agreements) or divergence (sum of disagreements) among stakeholders toward the same objectives. This can also help identify the restoration objectives that are gathering the most consensus and the stakeholders most involved in the achievement or nonachievement of these objectives. This can easily be read from the “correspondence map of stakeholders/objectives” (Fig. 4) in which close distances between different stakeholders and a given objective identify the level of agreement regarding the objective (Elmsalmi and Hachicha 2014).

The GDA is the most oriented toward both “conservation” and “sustainability” objectives under collective land tenure systems (see Fig. 4). This can be problematic since this stakeholder has been identified as completely dominated, which might affect sustainability of the restoration process. It is also important to note that technical public agencies have a conservative agenda that does not consider rangeland productivity. Thereby, rangeland productivity is the main concern of private rangeland users, and no restoration strategy will succeed without considering alternative feeding resources for farmers' compensation.

For the case of private rangelands, “conservation” and “productivity” objectives are consensual across GDA, beneficiary farmers, and OEP. Results also reveal consent around the “scaling up” objective, from CRDA/PRODESUD and OEP. The “scaling up” and “sustainability” objectives remain conflictual with a single concerned stakeholder for each objective.

The levels of power, potential alliances, and conflicts of all stakeholders involved in rangeland restoration under different land tenure systems are summarized in Table 3. For instance, these findings can be used to identify and also envisage strategic pathways to enhance the governance of rangeland restoration.

Discussion

Results from the analysis of direct and indirect influences among stakeholders involved in rangeland restoration, as well as the power relations, show that several important elements should be considered for collective restoration actions. Convergences and divergences of stakeholders regarding different objectives of rangeland restoration may affect the success of this process and, thus, need to be well identified and characterized under both private and collective contexts.

Stability and success of rangeland restoration processes in collective rangelands using the Gdel technique for a period of time would require some of the key stakeholders to increase their involvement, influence, and dominance. One of the main results concerns the need for better involvement of local government authorities in the Gdel process. This stakeholder has a favorable balance of power (strong influence in the rangeland restoration decision process) and at the same time is a relay stakeholder in the case of collective rangelands. Thus, their strong involvement through enhanced alliances with development and syndical stakeholders (CRDA, PRODESUD, and UTAP) can better stimulate the coordination of the restoration process (planning, implementation, and monitoring of rangeland restoration). This could also bridge the intervention gap between the two contexts (private and collective) and all the key concerned stakeholders, such as CG and GDA, considered as dominated and without an influential balance of power. This result is promising, considering the new trend in terms of political and institutional decentralization in Tunisia where communes (municipalities) are supposed to play larger roles in local development, including collective land management.

Table 3

Summary of relationships and power relations among the main stakeholders involved in rangeland management in Tataouine: potential alliances, conflicts, and strategic pathways for enhanced rangeland restoration.

(continued)

Fig. 4.

“Correspondence maps of stakeholders/objectives” in collective (left) and private (right) rangelands (Source: own elaboration from MACTOR analysis, 2019).

Besides, as Trumbull (2012) argued, it is not so much organizational issues that stand behind a successful collective action in public policy as it is legitimization. Various coalitions succeed when they are able to create a trustworthy in a transparency collective action arena, convincing the public and decision makers that their interest is compatible with general welfare and morally legitimate. Thus, the alleged weaknesses of latent groups seen in poor organizational abilities and differentiated interests may become their strength, because public opinion perceives their endeavors as more sincere than ambitions of organized and well-financed pressure groups. The former also tend to contribute to the general welfare instead of satisfying particular interests.

Furthermore, and given the lack of knowledge of other potential rehabilitation approaches such as reseeding grass species (Rinella et al. 2015), planting fodder trees (Guyassa et al. 2014), and improved grazing systems (Vallentine 2012), a research structure such as IRA should be much more involved in this rangeland restoration process. This could be effective through the valorization of its dominance and favorable balance of power to ensure its roles in scientific and technical support to all stakeholders. This can certainly help in implementing large rangeland restoration programs from technical, institutional, and organizational (setting better governance structure) perspectives in the region of Tataouine.

As discussed earlier in the "Introduction" section, the amount of information available to actors is one of action arena attributes that affect preferences, strategies, and actions. Collective action inherently involves a greater role for user groups in natural resource management but also regulating agents to transfer knowledge and information. This might include responsibility for monitoring resource use (by group members and outsiders), enforcing rules, and providing operation and maintenance services (Kiser and Ostrom 1982).

Other case studies have shown similar results in enhancing relationships between involved stakeholders rangeland management, such as the case study of rangelands of the western Great Plains of North America in the framework of "Collaborative Adaptive Rangeland Management" project aimed at fostering science-management partnerships and data-driven rangeland management through a participatory, multistakeholder approach (Wilmer et al. 2018).

In the case of private rangelands, rangeland restoration is much more stable, flexible, and successful since the majority of stakeholders have the same power in terms of influence and dependence but also several alliances among the majority of stakeholders and many more consensual objectives. In fact, people are more likely to abide by and enforce rules that they have had some say in setting. Thus, if collective action on rangelands is to give end users an equitable role in governance of the resource, they must be able to contribute collectively in setting rules, determining sanctions, and making critical decisions (Wheatland and Chêne 2015).

Findings show that development stakeholders (CRDA/PRODESUD and OEP) have strong alliances with research, CBOs, and local beneficiaries (which are in a favorable balance of power) and have consensus with them concerning objectives of conservation, productivity, and scaling up of the Gdel techniques. Some disagreements among stakeholders are recorded in regard to objectives of productivity and scaling up (e.g., between CRDA and CBOs). In this case, it is recommended to reinforce the balance of power of OEP, which is the most involved stakeholder in the collective restoration process. The process should be feasible by shifting to a relay between CRDA and CBOs to reach the objective of sustainability (conservation and productivity). The research structure (IRA) must also be heavily involved to ensure an enhanced role in scientific and technical support. In such a case, the priority is outlined according to landowners and end users (beneficiaries), CG, and authorities toward short-term benefits from the Gdel technique (productivity). Long-term objectives (conservation and sustainability) have been announced by development stakeholders (CRDA, PRODESUD, and OEP) and CBOs (GDA).

Through this study we try to give key pathways to reconcile “conservation” and “productivity” objectives and build a common front to reach the objective of rangeland sustainability, by managing relations between stakeholders, especially in the collective rangelands characterized by problems of access and use of common pool pastoral resources.

Furthermore, in the framework of the current transformation process in Tunisia from centralized toward integral and participatory management, managers should try to find options and actions (e.g., financial autonomy and capacity building) to improve the balance of power of the GDAs because they are completely dominated, and their participation is crucial for long-term success of a rehabilitation endeavor. Managing this transition requires a joint learning process and successful management of a multilevel governance model to support communication among different relevant stakeholders with a long-term focus (rangeland sustainability), including uncertainties and risk (e.g., climate change). This calls for concerted efforts, involving several stakeholders, both governmental and nongovernmental.

Strengthening of local organizations will enable local people to implement and sustain rehabilitation activities. Rehabilitation operations should also consider local peoples' short- and long-term needs and value systems in order to sustain their participation and interest. Trade-offs may be required to meet both the needs of these people and ecological objectives. This implies that more emphasis should be made on awareness raising in the communities on the importance of rangelands, particularly rangeland fencing techniques, social commitment, and strong cooperation among all local and regional stakeholders when choosing rehabilitation approaches. Furthermore, new policies should not only consider local perceptions of rested areas and effectively integrate the institutional roles and legislation in governing the collective and private rangelands but also address well-defined land tenure and/or secure property rights for these two rangeland categories.

Conclusion

This research provided pathways for enhancing collective action related to rangeland restoration programs in collective and private rangelands in Tataouine Governorate of southern Tunisia. We considered the richness and complexity of social pastoral systems and tried to understand the stakeholders' network and behavior, their mutual influences and dependencies, their balance of power, and the consensual and conflictual objectives—thus formulating key questions and strategic recommendations, which are important steps in the prospective process.

On the basis of the analysis of stakeholders' relationships in the two main land tenure systems of the study zone, this research offers potential synergies and compromises among the stakeholders in order to reach all assigned objectives (productivity, conservation, sustainability, scaling up, and adaptation to climate change), which are sometimes antagonistic (e.g., productivity vs. conservation). This helps to determine how these relationships might evolve in a participatory management (coordination, cooperation, and collaboration) and how particular stakeholders should value their dominance and others should improve their balance of power and be more involved in the rangeland restoration process.

The study showed that the Gdel technique is more effective in the private sector, given the multitude of alliances among stakeholders and consensus, especially on the objectives of conservation, productivity, and thus sustainability. The problems of collective rangelands are due to the conflictual relations between the development sector and CBOs on one hand and authorities on the other. The study also showed the importance of the role that authorities have to play in strengthening alliances and commitments with stakeholders that present conflicts of interest, especially where all “conservationist” stakeholders (CRDA, PRODESUD, OEP, and GDA) are dominated stakeholders while the “productivists” are dominant or relay (beneficiaries and authorities) except CG, which is dominated. This can affect the process of rangeland restoration in these areas and the achievement of assigned goals of good governance and sustainability.

The second outcome concerns the suggestion of key strategic recommendations. First, in the case of collective rangelands, stronger involvement of local territory development authorities in restoration process coordination could bridge the gaps between other stakeholders, such as to involve CG and GDA. Second, relevant stakeholders in the two land tenure systems could build a common front to reach both goals of productivity and sustainability. Third, in the framework of participatory management, managers should try to find options and actions (e.g., financial autonomy and capacity building) to improve the balance of power of the GDAs.

Stakeholder analysis with MACTOR showed its importance in the rangeland management process in general, and the Gdel technique in particular, as it helps in understanding and improving the collaborative modes that integrate local and institutional stakeholders. Despite these findings, the MACTOR method has some limitations, notably concerning the gathering of required input. Stakeholders are naturally reticent to reveal their strategic objectives and means of their external action.

We conclude generally that rehabilitation of degraded rangelands is challenging due to the complex nature of degradation drivers and the need for collective actions. Although it is based on the local scale, the case presented here gives some useful insights for institutional design of local-specific restoration approaches. This can also be applied more widely to rangeland management in other regions with similar contexts such as in the Mediterranean region.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.