Collaborative management in wolf licensed hunting: the dilemmas of public managers in moving collaboration forward

Although wolf recolonization can be considered a success in terms of population increase and geographical dispersal, the return of grey wolves Canis lupus to rural central Sweden has caused frustration and discontent among local stakeholders. Farmers and hunters living in — or adjacent to — wolf territories perceive the political decision to support wolf recovery as intruding on local lives and restricting opportunities for small-scale farming and hunting. They feel that decision-makers have left the consequences of wolf recovery policies unaddressed. To overcome the failure of previous policies and increase local consensus, in October 2009 the Swedish parliament passed a resolution concerning the introduction of licensed hunting on wolves as a measure expected to promote local acceptance and facilitate dialogue among different parties. In doing so, the Parliament delegated to the regional authorities the responsibility to organize, coordinate and implement licensed hunting in the administrative counties concerned. According to Swedish wolf hunting regulations, the regional authorities are to be involved not only in the achievement of primary policy goals, but are also expected to overcome antagonism and conflict through collaboration. Questioning the normative idea which suggests that public managers are expected simply to implement national legislation at the local level, this paper argues that managers create practices and establish routines that enable them to cope with problems related to the realization of collaborative management. Through a combination of participant observations and semi-structured interviews conducted with authorities and field-staff in four administrative counties during the implementation of licensed hunting, this paper concludes that in order to understand how collaborative management of natural resources works, greater attention has to be directed to the way public managers organize their activities, how they cope with their mandate and how they themselves relate to networks of actors.

Th e story of wolf recolonization in Sweden can be considered a successful one in terms of population increase and geographical dispersal. Although wolves were considered almost extinct before the enactment of the Preservation Act in the mid-1960s, the current population in Norway and Sweden is estimated at about 289 -325 individuals (Persson andSand 1998, Wabakken et al. 2011). Wolf establishment is widely regarded as incompatible with several confl icting activities, such as Sami reindeer herding in northern regions and sheep farming in central and southern regions. Hunters consider wolves competitors for game species, such as deer and moose. Additionally, wolf predation of hunting dogs has had negative reactions and further fueled controversy. Local discontent and frustration is amplifi ed by residents ' distrust of public agencies engaged to spread knowledge to the local community, increase acceptance and eff ectively implement policy (Cinque 2008, Sj ö lander-Lindqvist 2008, Sandstr ö m et al. 2009. Th e situation is exacerbated by lack of dialogue, as these two sides have divergent conceptions of how to preserve and maintain the current wolf population. Th e wolf is protected as an endangered species under international law (Bern Convention CETS 104) and EU law (Council Directive 92/43/EEC, Birds Directive 2009/47).
Th e Habitats Directive defi nes the wolf as an animal of " community interest in need of strict protection " and " prioritized " under Annex IV(a). Th is status means that Member States must follow a programme of strict protection for wolves, including prohibiting their intentional capture or killing. However, derogation is permitted in Article 16, to prevent serious damage, in particular to crops, livestock, forests, fi sheries and water and other types of property. Derogation is also permitted to preserve the interests of public health and public safety, for the purpose of research and education, and to allow, under strictly supervised conditions, on a selective basis and to a limited extent, the taking or keeping of certain specimens of the species listed in Annex IV in limited numbers specifi ed by the competent national authorities.
In Sweden, the regulatory framework for protection of wolves is expressed in the Environmental Code as well as hunting legislation (Hunting Act 1987:253, Hunting Ordinance 1987. According to Section 23c and 23d of the Hunting Ordinance, Swedish Environmental Protection Agency (SEPA) can decide on ' license hunting ' under those conditions mentioned at the start of Article 16 in the Habitats Directive; that no other satisfactory solutions exist and it is not detrimental to the maintenance of a favorable conservation status of the population in its natural range. A further condition is that such hunting is appropriate with regard to the size of the population and its composition. Hunting is also required to be selective and conducted under strictly controlled conditions (Darp ö and Epstein 2015). Based on these principles, in 2009 the Swedish Parliament decided to authorize wolf licensed hunting stating that SEPA should be responsible for setting up the hunting quota every year based on prediction models that take illegal hunting, authorized lethal killing, and the accidental killing of wolves in traffi c and other accidents into account. In those territories interested in hunting, County Administrative Board (CAB) managers and fi eld-staff became responsable for planning, monitoring and evaluating licensed hunting (SEPA Decision 2009-12-17). In particular, they were responsible for building, facilitating and supporting communications between the authorities and the hunting organizations, as well as to inform and communicate with communities living in hunting areas. Since the regulation states that licensed hunting should be adaptive, CAB managers acquired the main responsibility for designing and implementing the entire process.
Th is study investigates how public managers (frontline and fi eld-staff ) organize and carry out collaborative management of quota-regulated licensed wolf hunting. Th e study departs from the changing nature of the Swedish wolf management resulting from the policy decision to increase local participation and involvement through the introduction of quota-regulated wolf hunting (Government bill 2008/09:210:47). In this paper, I shall focus on the central role of public managers in organizing and carrying out collaborative management. More precisely, I shall investigate how they face the diff erent dilemmas that arise during the implementation of license hunting and what can be learned from this experience that could be used to improve future iterations of the hunt.

Wolf licensed hunting
In October 2009 the Swedish Parliament voted for a reform package that included a substantial reorganization of the management of protected large carnivores, consisting of decision decentralization and measures for the genetic vitalization of the wolf population (Government bill 2008/09:210). Furthermore, the new policy prescribed the introduction of 20 wolves from Finland and Russian Karelia in order to strengthen the genetic diversity of the population. It also confi rmed the previous decision that no wolves packs should be allowed within the all-year reindeer herding regions of northern Sweden. Th e cornerstone of the new legislation was the introduction of quota-regulated licensed wolf hunting implemented through a collaborative management approach indicated as the main strategy for organizing and conducting wolf hunting (SEPA 2010). Th e reasoning behind this decision was that collaborative licensed hunting was the only way to deal with social confl icts arising from the existence of wolves and to increase their acceptance in rural areas. It was said that to facilitate and support the coexistence of humans and large predators, people living in predatorinhabited areas had to be given better opportunities to infl uence decisions aff ecting local environments, as well as opportunities to take an active part in wildlife management (Sj ö lander-Lindqvist et al. 2010). According to this view, managers and fi eld-staff employed at the CABs were explicitly responsible for organizing and coordinating licensed hunting, increasing opportunities for participation and involvement through a collaborative approach.
Th e fi rst licensed hunt for wolves in Sweden in 45 years commenced in the early hours of 2 January 2010. According to SEPA, this wolf hunt attracted over 12 000 registrants. However, estimates from the Swedish Hunters ' Association indicate that only about 4000 hunters participated in the hunt. Shortly after the fi rst day of hunting, the wolf hunt was called off in two of the fi ve counties that had been selected as the geographical focus of the hunt: Dalarna and V ä rmland. In these counties, the assigned two-thirds of the national quota (nine wolves in each county) had been shot. After the third hunting day, the 4 January, a single animal remained of the quota of 27 wolves, with one overshooting in Dalarna. Out of the 28 animals shot, 46 percent were bitches, and most were relatively young individuals. All 28 wolves were found to be healthy and of a non-eastern origin after necropsies carried out by the Swedish National Veterinary Institute.
Th e current shift from governing and managing simple ecosystems to overseeing complex, adaptive, socio-ecological systems demands an integrated human-in-nature viewpoint for handling surprises and uncertainty. However, this shift creates a need for considering how diversity in individual and collective human responses relates to, and shapes, the systems (Newell et al. 2005, Folke 2007, Ostrom et al. 2009).

Collaborative management and leadership dilemmas
In recent years a worldwide intensifi cation of collaborative modes of work in public management has occurred (O'Toole 1988, Th omson 2002, Mandell and Steelman 2003, Huxham and Vangen 2005, McGuire 2006, O'Leary and Bingham 2009. Government incentives and directives relating to collaborative initiatives are increasingly abundant and many require the participation of local communities in order to achieve sustainable outcomes (Imperial 2005). Th is is due to the assumption that collaboration deals more eff ectively with major issues that cannot be tackled by the central state alone or solved easily by single organizations (McGuire 2006, Emerson et al. 2012. Several key components motivate the need for a collaborative management approach, sometimes deriving from a focus on effi ciency and sometimes from a concern to answer citizens ' expectations of equitable treatment, fi scal prudence and transparent process. Th e benefi ts of collaborative modes of work have been illustrated by several studies which also have underlined the diffi culty of achieving sustainable collaborative modes of work in practice Vangen 2004, Gray 2007).
In essence, collaborative management refers to a process involving diff erent actors with varying degrees of power and responsibility that interact and negotiate through formal and informal rules in order to develop mutually-agreeable solutions (Wood and Gray 1991, Huxham 2003, McGuire 2006, Th omson et al. 2007, Margerum 2008, Johnston et al. 2011). Since collaboration entails that several activities are negotiated and mediated (Ostrom 1998), it places great demands on managers ' ability to actually build, facilitate and support the process. At the same time, public managers are forced to play a leadership role in moving collaboration forward by harmonizing the multiple interactions of scale and level, maintaining administrative rules and achieving the specifi c targets of the policy decisions (Ansell and Gash 2007). Th ey have to cope with the innumerable relationships and interactions created by the collaborative management approach. Particularly in the fi eld of natural resource management, public managers become responsible for designing the process, including various stakeholders, and evaluating the outcomes (Conley and Moote 2003). Th is demands great ability in infl uencing and steering the collaborative process, leading and coordinating the participants, and empowering those who can deliver collaboration outcomes (Huxham 2000). At the same time, as the collaborative process is highly decentralized and inclusive, the leadership function should be largely facilitative, supportive and never manipulative (Feldman and Khademian 2000). A leader should steer the process without directing the course of the process and they should empower a broad membership without favoring any particular interest.
Studying how leadership might impinge upon the practice of collaboration, scholars fi nd that managers cope with a state of uncertainty in making decisions (Susskind and Cruikshank 1987, Gray 1989, Edelenbos 2005. For example, Vangen and Huxham fi nd that the more that stakeholders fundamentally distrust each other, the more leadership must assume the role of honest broker. However, when incentives to participate are weak or when power is asymmetrical, the leader must often intervene to help keep stakeholders at the table or empower weaker actors. Th ese diff erent roles may create tensions because intervention to empower weaker actors may upset the perception that the leader is an honest broker (Vangen and Huxham 2003). Based on the experience of previous research, I framed three main dilemmas that public managers need to cope with in order to move the collaborative process forward.
Th e fi rst dilemma (the dilemma of democracy versus effi ciency) refers to the tension between the ideals that inspire collaborative processes (as for example the inclusion of a broad representation of interests or the adoption of a deliberative decision-making process) and the need to quickly accomplish collaborative goals. In a more idea-oriented leadership style, the managers allow all members to have a voice and eff orts are directed at facilitating the involvement of any individual or organization that is willing and able to improve the collaborative agenda. Possibly, one of the biggest challenges of following the collaborative ideal is to fi nd ways to address the principle of ' community involvement ' that is frequently deemed an important aspect of collaborative governance. On the other hand, when managers adopt a pragmatic way to manage, they risk to impose their own interpretation and understanding of problems under debate.
Th e second dilemma (dilemma of autonomy versus control) refers to the degree of control and authority that should steer the process. Mostly at the implementation phase, leaders need to make a decision to secure their broad-based infl uence and control, or stimulate creativity by providing autonomy to the diverse participants, so that the group can generate new ideas and perspectives. In identifying this tension, Yan and Gray (1994) observe that there is a positive link between level of control and level of trust; managers are concerned and only willing to relax control in situations where trust is high. Th is present a real challenge to those considering to set up collaboration where there is no history of relationship between the participating organizations or where previous relationships have not generated mutual trust (Huxham and Vangen 2005).
Th e third dilemma refers to the evaluation stage. Th e managers evaluate the outcomes and either decide to maintain the process for the future (and eventually institutionalize it) or return to the previous situation (the dilemma of stability versus change). A comprehensive and impartial evaluation of the benefi ts is clearly crucial for this stage to properly inform future decisions.
Scholars do not agree on the most successful way in which public managers should resolve those dilemmas. As noted, the challenges of managing the collaborative process vary greatly depending on the particular case (Vangen and Huxham 2003) and the local context (Ansell and Gash 2007). Inevitably, therefore, guidance for practice cannot be precise. Even so, in this paper, I shall provide a tool for considering and addressing the challenge of developing collaborative management, from the perspective of the actors whose role it is to move collaboration forward: public managers. Similarly to previous studies of collaborative governance that describe collaboration as developing in stages (Susskind and Cruikshank 1987, Gray 1989, Edelenbos 2005, I have organized this paper to correspond to three conceptual collaborative realms that emerged from the diff erent phases of the organization of wolf hunting. Th ese are: 1) pre-phase; 2) licensed hunting; 3) post-phase. Each phase corresponds to one of the actions or activities that public managers engage in to cope with, or build upon, the constraints or possibilities dictated by the collaborative process (Fig. 1).

Material and methods
Th is study investigates how public managers (frontline and fi eld-staff ) organize and carry out collaborative management of quota-regulated licensed wolf hunting. Th e data were collected from December 2009 to March 2010 in four of the fi ve Swedish counties where licensed hunting was fi rst carried out. Th ese counties are: V ä stra G ö talands, Dalarna, V ä rmland and Ö rebro (Appendix 1). Th e data was derived from a) eight participant observations, b) 16 face-to-face interviews and c) 12 telephone interviews. Interviews were spread equally over the four administrative counties. on collaborative management in order to draw the results. A two-step process outlined by Hay (2005) has been employed in order to distinguish overall themes, followed by a more in depth, interpretive code in which more specifi c trends and patterns can be interpreted. Th is process involves a circular hermeneutic movement.

Results
In this section, I shall report the results of the interviews and participant observations conducted in four administrative counties. As previously mentioned, I have organized the section to correspond to the three conceptual collaborative realms that emerged from three diff erent phases in which wolf hunting was organized: 1) pre-phase; 2) license hunting; 3) post-phase. In each phase, I focus on issues raised by subjects and identifi ed as signifi cant. Each of the three subsections below, which are example-based rather than comprehensive, describes a category of activities that was repeatedly observed and reported during data collection.

Pre-phase
Although in most counties the wolf hunt was called off after only a few days, planning for licensed hunting had been intensely undertaken by the hunters ' associations and local authorities in the months preceding the event. Th is stage, the pre-phase, refers to the period of preparation for wolf hunting when a number of information meetings and study circles were arranged by the CABs and the Swedish Hunters ' Association to inform and educate hunters intending to participate. Interviews conducted with both frontline and fi eld-staff reveal that the pre-phase was characterized by attempts to build trust and foster a feeling of inclusion. At the same time, the interviews clearly highlight a tension between the ideal that trust and inclusion are needed for collaboration to be successful and the requirement for pragmatic steering of the process in order to achieve the goals. More precisely, managers struggled with the challenge of how to create personal dispositions to trust when institutionalized practice of mistrust have dominated the scene for a long time.
In V ä rmland and Dalarna, two administrative counties where the confl ict between CABs and hunters was traditionally strong, managers established mechanisms to control behavior in order to increase the performance of the process. For example, the information meetings organized by the Swedish Hunters ' Association before the hunting were declared as mandatory for taking part to the wolf hunting. In those meetings, the CABs presented the Government decision to support hunting as a promising approach to solve the confl ict aff ecting the management of large predators in Sweden. At the same time, it was stated that hunting was a temporary event which could be revoked by the Government if rules were broken. Th e CABs vigorously addressed the importance of organizing the hunt in a professional and trustworthy way, encouraging hunters to obtain as much information as possible on wolf behavior, stamina and adaptability in diff erent situations.
Another strategy to create consent was to recognize past failures in the implementation of wolf policy. For example, at a public lecture organized by the hunters ' association However, observations were not conducted in the county of Ö rebro.
Data collection was carried out in the three separate stages of licensed hunting. Th e fi rst phase, which corresponds to the pre-phase of licensed hunting, included planning activities, public information meetings and local study circles. At this stage the research team carried out participant observations and semi-structured interviews with frontline CAB managers and fi eld-staff . Th e second stage concerned the actual licensed hunting. During this phase public managers continuously communicated with the hunters and mobilized them, thus essentially controlling the hunt. In this stage the research team carried out telephone interviews with frontline CAB managers. Th e last stage, which relates to the postphase of licensed hunting, included process and outcome evaluation. In this stage, telephone interviews were carried out with CAB managers and representatives from the Swedish Hunters Association to elicit their thoughts and perspectives on the recently concluded licensed hunting.
Th e face-to-face semi-structured interviews were organized according to a semi-structured interview plan and they were carried out with frontline CAB managers and fi eld-staff . Th e plan allowed the interview subjects to expand on themes that they considered to be of particular interest. Th e original themes of the interview plan covered diff erent aspects of preparing the licensed hunt. Managers (both frontline and fi eldstaff ) were asked to describe their expectations and beliefs, for example, about their own actions during the planning and the organization. Th ey were also asked to describe their relations with the other CABs and with SEPA. A standard note-taking procedure was employed during the interviews.
Telephone interviews were carried out with frontline CAB managers and they were structured according to an interview plan consisting of questions about how they chose to organize collaborative wolf hunting, as well as associated problems and considerations for the future. Th e plan also included questions about their own activities during the wolf hunting and " lessons learned " . Th e telephone interviews were recorded and transcribed and the results coded and analyzed.
Participant observations were conducted at meetings and local courses in the pre-phase of licensed hunting. Th is was done to enhance the data as it gives the researcher a greater sense of the social status of the people and activities studied (Dewalt et al. 1998). Th e researcher introduced herself at local courses as an observer, explaining that her observations would lend further insight into the public involvement process and making clear the topic of the research project.
Both interview data and observation notes were, as far as possible, coded and analyzed with respect to fi eld-specifi c perceptions and concepts. For example, observations were reconstructed and analyzed, enabling the researcher to discern and interpret the main points of debates and discussions, i.e. themes. Th emes emerged based on the emphasis informants laid on a particular issue or phenomenon; how issues raised by informants were related to the objectives and research questions. Also, informants ' narratives were transcribed, revised and examined in light of their responses, to allow an analysis of tacit processes, ideologies and power relations (Fangen 2005). Subsequently, themes have been analyzed in the light of existing theory and previous research We should collaborate with those two categories of interests as we implement wolf hunting " .
Later, during the interview, the same informant explained that the real meaning of the word " collaboration " is not to achieve a consensus on wolf management, but rather to " openly recognize the fact that hunters and farmers are key players in the survival of local communities " .

Implementation phase: licensed hunting
Th e fi rst licensed hunting in Sweden commenced on 2 January in the fi ve administrative counties of Dalarna, G ä vleborg, V ä rmland, V ä stra G ö taland and Ö rebro. Each county received its own quota to fi ll. SEPA approved the culling of 27 wolves, but 28 were actually killed. Th e decision meant that anyone with hunting rights in those areas had a right to participate, providing he or she registered a few days before hunting started. To fulfi ll the requirements of a controlled hunt with clear accountability, the authorities decided that hunting would not be conducted without a responsible hunt leader. His or her task was to ensure that offi cial information on the current hunting status reached the hunters on the fi eld. At the same time, each hunter was instructed to immediately report to the authorities every culled animal. According to the regulations issued by SEPA, the hunt leader would report the shooting location and number of wolves killed to the fi eld-staff as soon as possible by mobile phone.
Th e semi-structured telephone interviews highlight a sense of uncertainty associated with the question of how much direction should be given during wolf hunting. Th e CABs representative referred to this as " the dilemma of choosing between control versus autonomy " .
Th ey thought that too strong a leadership would have negatively aff ected collaboration. Th us, they preferred to delegate signifi cant responsibility to fi eld-staff during the wolf hunt. One CAB staff member justifi ed this decision with the consideration that, from hunters ' perspective, CAB fi eld-staff is viewed as more neutral and friendly than managerial staff . In the interviews, they referred to this strategy as " intermediate management, " arguing that " when the level of confl ict is high, it is more eff ective to delegate to third-party negotiators " .
Although the use of intermediate managers was presented as a viable compromise between autonomy and control, its own existence implied that the authorities only partially trusted the hunters. Th is was confi rmed by the interviews conducted with the managers, who also mentioned the diffi culty of picking up fi eld-staff who are not allied with the hunters but can still give them the impression of being on their side. As a Dalarna CAB manager remarked during the interview, the practice of creating a buff er between the hunters and the CABs is a necessity rather than an indication of trust: " Th e delegation of authority to fi eld-staff and their presence in the specifi c areas appears as an attempt to overcome precedent confl icts, as a sign of trust, so to speak. In reality it is inevitable because we cannot be everywhere at the same time " .

Post-phase
Th e post-phase regards the evaluation stage that occurred after wolf hunting was concluded. During this stage CAB to inform fellow hunters, the invited CAB representative openly recognized the local administration ' s past failures in terms of lack of communication with local communities. At another lecture, a member of the fi eld-staff operating in V ä stra G ö taland County also expressed his frustration at not being able to communicate with people directly in the past, having been required to delegate communication to the interest organizations. A member of the fi eld-staff of the Dalarna CAB strongly expressed the view that licensed hunting could represent a new start for the parties involved, emphasizing that this represented a fi rst step in involving hunters in the management of large carnivores.
Th e task of keeping a balance between recognizing mistakes and improve changes was also stressed during the interviews as a trust-building strategy. When talking about the hunt-preparation stage, for example, several informants pointed out the necessity of promoting a sense of transparency and openness, because wolf management in Sweden has long been hampered by a low level of trust. However, informants pointed out that their participation in the meetings also was motivated by other concerns, such as the need to keep the discussions under control or to see what kind of information was being provided at the meetings themselves (i.e. brochures, fl yers and books).
Closely linked to the above behavior was the attempt to create an identity for those involved in the process, in order to foster a feeling of inclusion. Th is was addressed explicitly at the meetings and also confi rmed during subsequent interviews. At the meetings, CAB representatives were keen to acknowledge that the hunters collectively brought a range of experiences, skills and perspectives to the collaborative process. For example, a representative from the Ö rebro CAB referred to hunters as " pillars of local communities, " referring to their role in keeping the Swedish hunt tradition alive. At one public lecture, a representative from the Dalarna CAB recalled the hunters ' role in helping the authorities during the annual wolf tracking that is carried out to collect information about the number of wolves and packs populating the diff erent areas. He also referred to social and cultural aspects of wolf hunting pointing out that " helping farmers and landowners to prevent damages caused by large carnivores has an important symbolic meaning for local residents. " When the meetings ended, I interviewed the CAB personnel, asking them to confi rm or deny whether they had intentionally tried to create a sense of collective identity by creating a picture of Swedish hunters as engaged and responsible individuals. All interviewed CAB staff confi rmed this intention explicitly, and clarifi ed that it was driven by the necessity to create a connection between hunters and management. One explained that a major problem in current wolf regulations in Sweden is the fact that it tends in practice to create a separation between farmers or hunters and CAB staff . Th is separation is formally justifi ed by the necessity of fair and impartial public administration. However, the situation is far from been clear-cut, because most CAB staff working with large carnivores are also hunters or have a strong connection with local farming activities. A Dalarna CAB frontline staff member stated: " I grew up in a hunting family: my parents were hunters. But I also grew up on a farm. So, it is quite natural for me to understand both farmers and hunters when they complain about large predators.
is whether or not managers should communicate to SEPA regarding mistakes occurring during the implementation stage, in order to introduce modifi cations and adjustments for the future. Th is is described by many CAB staff members as a risky decision, as the open recognition of mistakes and faults can be employed by skeptical parties to help them criticize collaboration and question the legitimacy of the process. For example, as both managers and fi eld-staff reported, the atmosphere during the hunt was perceived as stressful as there were many hunting teams that wanted to kill a wolf before the quota was full. Some interviewees also refer to hunters shooting wolves without much control, and some wolves were injured before being killed. However, despite their perceiving the situation in this way, these CAB representatives said they decided not to report these facts to SEPA in order to avoid negative consequences on the possibility of allowing future hunts.
According to a V ä rmland CAB manager, the decision not to report negative events that occurred during the hunt derives from the ambiguity of the system. In other words, it would be a consequence of the fact that managers are at the same time controllers and controlled: "Th e law requires us to carry out an assessment without taking into account the fact that our judgment shall aff ect our future work, even in a negative way. We are both controllers and controlled, and these two confl icting roles cannot agree".
Similarly to the case of the inspections of damage on living private property caused by large carnivores (Sj ö lander-Lindqvist and Cinque 2013), the Swedish legislation on wolf hunting creates a role ambiguity whose solution is entirely left to the CABs.

Discussion and conclusions
Th is paper provides a perspective on collaborative management focusing on how managers organize and coordinate licensed hunting. In making this attempt I have taken into account their behavior, underlying the strategies chosen to achieve collaboration and cope with the dilemmas that arise during the diff erent phases of the process. As a result of this approach, the study includes some lessons for public managers facing the challenges and opportunities of collaboration.
Th is paper shows that managers actively work to generate trust and move beyond the dissatisfaction of community members with past government eff orts. Th is is particularly true during the initial stage or pre-phase of the collaborative process. Th is result is in line with the results of other studies on collaborative management that also point out that those leading the process should be able to overcome the feeling of antagonism and mistrust among the participants before they initiate the process itself (Susskind andCruikshank 1987, Ansell andGash 2007). Here, memory and sense of identity matter, and since histories can be quite painful and lasting, managers cannot ask citizens to " check their pain at the door " (Forester 2006). Because of this, reconciliation between hunters and managers remains problematic. For example, the decision of CABs to make use of fi eld-staff in order to monitor the hunting suggests that their trust in the hunters is only partial. At the same time, managers are very concerned about the risk of inspiring critics or giving managers and hunters met each other to draw results and evaluate outcomes. Th us, this phase is an essential pre-requisite to address the third dilemma ' stability versus change ' .
Most of the informants began the interview by pointing out that the authorities had a very short time -one and a half months -to organize the activities related to the hunt. Most informants said that the decision to hold the hunt came too close to its implementation. And in fact, both CAB management staff and fi eld-staff explained that they had tried to fi nd some strategy of their own to cope with the defective rules and guidelines. For example, all counties provided a dedicated phone number at which hunters could obtain information on how many wolves had been eliminated in each area and how many remained to trap. A V ä rmland CAB staff member got the idea to write a blog reporting in real time what was happening out in the fi eld " … so even ordinary people could easily follow the hunt. " Both strategies were invented in the context of a bottom -up process where local authorities consulted the hunt leaders in each area asking for further cooperation and suggesting temporary solutions to cope with the short turnaround. As one CAB staff member explained, the result of such cooperation was the creation of closer and more open communication between groups that previously did not talk with each other.
According to the interviews, the CAB managers believe that the quota-regulated wolf hunt promoted broader participation in predator management, increasing public input. Th e public and open character of the 2010 wolf hunt, in which anyone holding a hunting permit could participate, is regarded by the informants as a signifi cant symbol of local justice and fairness. Informants agree that the politicians, by deciding to regulate the wolf population through licensed hunting, demonstrated that they cared about the community. Th is was an important message to convey to SEPA when the agency evaluated the hunt.
Also the telephone interviews with representative from the Swedish Hunters Association indicate that they interpret the event as a starting point for a long-term collaborative relationship with the CABs. For example, several informants say that after the hunting, they found that their relationship with the managers seemed to have been strengthened. From their perspective, the administrations credibility was increased. One informant explained that, whereas in the past there had been considerable distance between hunters and managers, after the hunting " the gap between the two parties has considerably diminished. " Another representative from the Swedish Hunters Association pointed out that the participants thought that the hunt generated a more positive attitude towards the wolf and its presence in the landscape. Wolves have been regarded by hunters and farmers as an animal with a special position in the Swedish fauna due to politics and management implementation. " By managing wolves like other game, such as elk or foxes, the wolf, according to several hunters, comes to be perceived as any other animal " . Th us, the decision to allow wolf hunting allowed the re-naturalization of a previously politicized species.
Despite the fact that the CAB staff share a positive and optimistic feeling regarding collaboration, they also say that the important question that must be resolved in the future suggestions to the SEPA, particularly during the evaluation phase. Considering that wolf hunting is a highly controversial topic for a majority of the population, these managers prefer to not draw attention to any and every error, in order to preserve the stability of the collaborative process. Th is behavior indicates that managers do not trust SEPA to truly embrace the idea that such a process is inevitably adaptive and must thus learn from this fi rst implementation of the hunt.
Facilitative leaders or intermediary managers are widely thought to positively impact the management of collaborative eff orts, by bringing together partners and making the various interests involved work towards the same goals (Ansell and Gash 2007). However, as this study demonstrates, the establishment of a buff er between hunters and managers is a necessity rather than a sign of trust from the administration. Th is refl ects the uncertainty that surrounds the practice of collaboration whenever existing confl icts may burden the process. It can be concluded that, when managers experience a low level of trust during the implementation phase, they employ intermediary staff as a control rather than a solution to the dynamic tension between autonomy and control.
In line with the results from another study on damage compensation caused by large carnivores (Sj ö lander-Lindqvist and Cinque 2013), this study also highlights the fact that lack of a more fl exible regulation generates role ambiguity and creates a quandary in the demand for neutrality in decision-making situations. In evaluating the process, CAB managers experience a dilemma generated by the fact that they constitute an integral part of the process itself.
In conclusion, the study highlights the transformation of the classical command-and-control model of management. Th is model is based on the assumption that public policy goals can be specifi ed clearly, and that the best way to go about achieving those goals is to use centralized authority to manage individual behavior. However, the rise of public participation complicates this vision, making it clear that the goals and scope of politicians do not always correspond to the public will. Preferences change for exogenous and endogenous reasons, and our understanding of public problems can change as diff erent evidence, language, and ways of examining the problem are brought to bear (Gusfi eld 1981, Stone 2000. In such a context, managers seem to play a key role in leading democracy forward, adapting rules and regulations to individual cases, facilitating dialogue between diff erent parties and mediating among diff erent interests, as well as advancing collaborative processes.