Langle-Flores, A.; Ocelík, P., and Pérez-Maqueo, O., 2017. The role of social networks in the sustainability transformation of Cabo Pulmo: A multiplex perspective. In: Martinez, M.L.; Taramelli, A., and Silva, R. (eds.), Coastal Resilience: Exploring the Many Challenges from Different Viewpoints. Journal of Coastal Research, Special Issue No. 77, pp. 134–142. Coconut Creek (Florida), ISSN 0749-0208.
Coastal tourism is often caught in a crossfire between economic benefits, ecological impacts and social tensions. Development of large-scale resorts can reduce the provision of certain ecosystem services and threat local people's livelihoods. Social networks might influence the transitions of governance systems into new adaptive models. We focus on the role of multiplex networks in the process of sustainability transformation by examining social networks that protected a marine reserve against the construction of a large scale development. The multiplex network exhibited a structure with five blocks: “scale-crossing brokers”, “visible leaders”, “ecosystem managers”, “visionaries” and “public sector”. This last block was structurally isolated from the rest of organizations. Multiplex networks facilitated the coordinated mobilization of information and resources across spatial scales. “Scale-crossing brokers” with the aid of “visible leaders” propelled up the local conflict toward national and global arenas, affecting the decision of Mexico's federal government to annul large scale resort's construction. Understanding the social processes that enable adaptive governance systems is crucial for sustainability transformations and resilience of coastal ecosystems.
One of the biggest challenges of humanity is to transform unsustainable social-ecological interactions into new trajectories that sustain the capacity of ecosystems to provide wellbeing for present and future generations (Moore et al., 2014; Westley et al., 2011). As part of this transformation, it is desirable to create resilient social-ecological systems (SESs) with capacity to absorb recurrent disturbances, and to retain essential structures, processes and feedbacks (Adger et al., 2005). Adaptive capacity, innovation, learning and transformability are essential dimensions of resilient social-ecological systems (Folke, 2006). Thus, desired adaptive governance systems must rely on institutions that encourage these aspects, integrate different types of knowledge in decision-making and promote collaboration across multiple scales (Folke et al., 2005).
Social network analysis has been widely used to understand these dimensions of social-ecological resilience (Bodin, Crona, and Ernstson, 2006). Each aspect has been linked to different functions and structures of social networks. For example, it has been argued that social networks play a critical role in the challenge to transform governance systems into new adaptive models (Folke et al., 2005; Moore et al., 2014; Olsson, Galaz, and Boonstra, 2014; Westley et al., 2013).
As Bodin and Prell (2011) argue, since SESs emerge out of complex interdependencies between as well as within natural and social environments, and our understanding of how we can sustain resilient SESs needs to embrace a relational perspective (see Abbott, 2007; Schneider, 2015). This approach seeks an explicitly processual understanding in which outcomes, actors, and relations are endogenous (Abbott, 2007). We subscribe to this position and address the role of social networks in governance system related to a large scale touristic development planned near Cabo Pulmo National Park (CPNP) in Mexico. We focus on the role of social networks in the process of sustainability transformation for CPNP and its implications in the resilience of coastal areas.
By using social network analysis framework (Wasserman and Faust, 1994), we examined the links between structure of social networks and sustainability transformations. We argue that multiplex social relations facilitated coordination of the mobilization of information and resources across heterogeneous organizations and spatial scales.
According to resilience theory, sustainability transformations consist of four phases: disruption, preparation for change, transition and institutionalization of the new trajectory (Moore et al., 2014; Olsson, Galaz, and Boonstra, 2014). In environmental governance such episodes are shaped by political interactions where state as well as non-state coalitions (with different interest and beliefs) share information and extend cooperation, bargain or conflict relations (Voß and Bornemann, 2011). Political change achieved by collective actions in policy processes can trigger the conditions for transformability of SESs (Ernstson, Sörlin, and Elmqvist, 2008; Moore et al., 2014). Transformative collective actions aim to disrupt governance systems through mobilizing social networks of support across scales and/or building collective action frames (Moore and Westley, 2011; Olsson, Galaz, and Boonstra, 2014). For example, Ernstson, Sörlin, and Elmqvist (2008) analyzed a collective action that institutionalized conservation efforts across various landscapes, resulting in an archipelago of remnant coastal forests into a single national urban park in Stockholm, Sweden.
In relational perspective, sociopolitical processes are seen as structured patterns of interactions among a variety of actors (Marsh and Rhodes, 1992) that, in the case of SESs or its subsystems, stretch across levels (Folke et al., 2005). These relational patterns can be captured by network measurements such as density or centralization. For example, whereas highly centralized networks are effective for simple and easily identifiable tasks (Borgatti and Everett, 2000), complex modular structures enable coordination across different levels (Girvan and Newman, 2002). Local measures, such as degree or betweenness centrality, identify organizations that occupy different central positions in the network (Wasserman and Faust, 1994).
The Case of Cabo Pulmo
Cabo Pulmo National Park is a marine reserve with an area of 71 km2, located near the southern tip of Baja California Peninsula, Mexico (Figure 1). The park was designated in 2008 as a wetland of international importance under the Ramsar Convention and belongs to the World Heritage list. One third of the park is officially a restricted area for fishing (25 km2), although local ex-fishermen embraced a no fishing policy all over the park since 1995. As a result, Cabo Pulmo National Park is considered an effective marine reserve that in 2011, generated US $590,400.00 through recreational diving and snorkeling activities (Reyes-Bonilla et al., 2014). CPNP and the adjacent village of Cabo Pulmo are under the economic and political influence of Los Cabos, one of the most important tourist destinations of Mexico, located 120 km southwest. Although, the expansion of tourism in the region has been the driver of large economic benefits, there have been also high social and environmental costs in Baja California Peninsula (Gámez and Ganster, 2012). The tiny rural settlement of Cabo Pulmo's is a highly contested space, with internal borders between the residential developments of foreign residents, mainly from the U.S. and Canada, and the modest houses without basic services of the Mexican residents. The village of 200 inhabitants faces internal divisions due to historical land transactions and ambiguous ownership rights to land (Anderson, 2014).
In February of 2008, a real estate developer applied to Mexican Ministry of Environment for a permit to build a resort of 3,814.6 hectares eleven kilometers away from Cabo Pulmo National Park. With an investment of 2.5 billion dollars, the resort would have created the equivalent to 30,354 housing units, two golf courses, and a marina (CAPSA, 2009). It was considered that the development of coastal infrastructure threatened the SESs by increasing the vulnerabilities of coastal areas, fading away the provision of certain ecosystem services and reducing the wellbeing of local communities (Adger et al., 2005). This situation that elicited the mobilization of an international coalition of organizations called “Cabo Pulmo Vivo”. The coalition reframed the protective history of Cabo Pulmo and launched a campaign named “Save Cabo Pulmo” that during the forthcoming years transcended from local to global arenas (Anderson, 2014). On April 2008, organizations showed their concerns in the public consultation meeting of the project: the potential effects of golf course runoff on coral reef, an increased sediment loading due to the construction of marina and breakwaters and the depletion of ground-water aquifers in the region. However, on September 2008, the first environmental impact authorization of the resort was issued (SEMARNAT, 2011).
Between 2009 and 2011, national NGOs claimed five reconsideration requests to the Ministry of Environment alleging violations to ecological planning regulations of Los Cabos municipality, in particular the construction on sand dunes. In August 2010, an administrative court took into account one of the claims and revoked the first permit. Nevertheless, in January 2011, the Minister of Environment issued a second partial permit excluding the marina and the water desalination plant. The permit was on condition that the developer conducted longitudinal studies on direction of marine currents (SEMARNAT, 2011). In May 2011, members of ex-fishermen organizations, national and international NGOs traveled to Spain to campaign against the investment bank that financed the resort's construction. Shortly after, the investment bank was declared in bankruptcy and nationalized. In June 2011, coalition's members performed a public demonstration in the national park, launching a giant lifesaving, 15 meters of diameter wide, with the legend: “Save Cabo Pulmo”. In reprisal, organizations were accused of damages to the coral reef by the federal government. Simultaneously, the developer released a social media campaign where they rejected and reframed the claims of environmental organizations and pointed to wrongdoings in the management of CPNP.
On November of 2011, a joint mission of RAMSAR, UICN and UNESCO visited the national park and met all the parts in the conflict. The joint mission recommended to include CPNP under Montreux Record as wetland site where changes are likely to occur by human interference. At the same time an international NGO released electronic communications where coordinated actions between federal government agencies and developer were revealed. In March 2012, an international NGO released a national campaign with prominent musicians and artists in order to gather online petitions for Cabo Pulmo. At the end of the month, activists were momentarily detained after displaying a giant banner in Mexico City appealing to the president to annul the permit. Two days later, the Minister of Environment appeared in a public hearing to the Senate Committee on Environment to respond on accusations about the resort's permits. On 3 June 2012, as part of a public demonstration in Mexico City's main square, members of the coalition delivered 220,000 online petitions requesting the project's cancellation to the federal government. Finally, on 15 June 2012, the president annulled the environmental permit alleging uncertainty to irreversible damages on the coral reef by the construction of the resort.
In the aftermath of this first episode of mobilization, coalition and their allies had successfully stopped in two occasions two slightly different versions of the same coastal development. Recently another coastal project planned in the vicinity of CPNP was rejected by the environmental authority. Crises on coastal planning like this one bring profound shifts that modify the human-environmental interactions and its feedbacks (in the sense of Moore et al., 2014; Olsson, Galaz, and Boonstra, 2014; Westley et al., 2013). As such, and if successful, they can be seen as political opportunity structures for social self-organization of governance systems toward sustainable transformations (Folke et al., 2005). In this context, Cabo Pulmo represents a critical case to study transformations in SESs where networks can build multiactor and multilayer coalitions in complex governance systems (Moore et al., 2014).
From October 2013 to April 2014, we approached 40 organizations involved in Cabo Pulmo's environmental conflict. From these, twenty-nine organizations accepted to be interviewed. With one exception, organizations affiliated with the resort developers declined to be interviewed. Thus, 25 respondent organizations of our sample were against the resort, three respondents were neutral and only one respondent (the developer) was in favor.
We conducted a face to face or video conference survey with 29 organizations (respondents). The 29 respondents (egos) reported 95 organizations (targets) with whom they extended a variety of ties through a data gathering technique known as a name generator (Lazega and Pattison, 1999). There was no limit in the number of targets that each respondent gave. We included only the initial set of 29 organizations that were identified as key members of the network. This list was consulted and validated via two expert interviews. In addition, based on the information from the survey, and after reviewing the website of each organization we differentiated branches from headquarters. In order to protect the identity of organizations due the episodic nature of the Cabo Pulmo's conflict and safety concerns, we built a four letter code to anonymize each organization and differentiate branches from headquarters, for example “I6B1” means “branch one of international nongovernmental organization number six”.
We asked eight different types of network questions in the survey: (1) Indicate a list of organizations with which your organization had any relationship from 2008 to 2012 regarding the conflict between Cabo Pulmo and the large scale development′s environmental permit. (2) From the list of organizations you mentioned, with which did your organization exchange ideas and information regarding the conflict? (3) From the list of organizations you mentioned, indicate those organizations with which your organization had work ties associated with the conflict. (4) From the list of organizations you mentioned, indicate those from which your organization received funds on activities related to the conflict. (5) From the list of organizations you mentioned, indicate those with which your organization had friendship ties. (6) From the list of organizations you mentioned, indicate those with which your organization shares the same political affinity. (7) From the list of organizations you mentioned, indicate those with which your organization share family ties. (8) From the list of organizations you mentioned, indicate those that stand out as most influential during the conflict.
We used relationship ties as proxy of cooperation linkages. Additionally, we applied to all respondents a semi-structured interview to elicit respondent's experiences and perspectives on the conflict (Klandermans and Staggenborg, 2002). We extracted from the interviewee's narration a network of organizations with whom their organization had conflict relationships. Based on the classification scheme proposed by Moeliono et al. (2014), we classified respondents and non-respondent alters organizations into ten types: (1) national non-governmental organizations, (2) international non-governmental organization, (3) donors, (4) universities and national research centers, (5) federal government agencies, (6) ex-fishermen organizations, (7) intergovernmental treaties, (8) companies, (9) media and (10) others. Ex-fishermen used to fish for a living, but after the declaration of the national park in 1995 they gradually turned into ecotourism providers.
For directed networks, density is given by the number of ties in a network divided by the number of all possible ties: N * (N − 1) where N is a number of nodes in the network (Wasserman and Faust, 1994). Thus, density ranges between 0 where there are no ties between any of the nodes and 1 where every node is linked to all other nodes in the network. Krackhardt's connectedness is defined as 1 − [V / N * (N − 1) / 2], which is the total number of pairs that are not mutually reachable (V) divided by the maximum number of possible pair combinations: N * (N − 1) / 2 (Krackhardt, 1994). Reciprocity is a proportion of mutual dyads (2 * M) to a sum of the mutual and asymmetrical dyads (A); i.e. 2 * M / (2 * M + A). Likewise, transitivity is given as a ratio of transitive triads to a total number of potentially transitive triads (Butts, 2008a,b). Also these measures range between 0 for networks without any mutual dyads, respectively transitive triads and 1 for networks where all non-null dyads are reciprocated, respectively all non-null triads are transitive.
Further, we calculated centralization scores for each network based on two basic centrality indicators: degree and betweenness centrality. Centralization is then given by the sum of differences of the highest centrality value from all centrality values in the network (Wasserman and Faust, 1994). The degree of centralization in a network is an indicator of hierarchy within the structure. It refers to the extent a network is dominated by a single actor. A maximally centralized network looks like a star where the node is at the center of the network with ties to all other actors (Freeman, 1978). Degree centrality is the sum of inward and outgoing ties to/from other organizations (Wasserman and Faust, 1994). Betweenness centrality measures the extent to which an organization lies along the shortest path connecting all other organizations in the network (Freeman, 1978). This variable refers to the proportion of all paths linking organization j and k passing through organization i. The betweenness of organization i equals the sum of all paths jk. Therefore, betweenness centrality is a measure of the number of times an organization occurs on geodesic paths between other organizations. The measure considers the positions of an organization's links such that the organization is more central if it is linked to more organizations that have many links themselves.
We adopted a multiplex perspective, which means that we have analyzed more than one type of tie between the same set of nodes (Heaney, 2014; Kivelä et al., 2014). Thus, multiplex networks occur as result of overlapping roles, actions, and affiliations in a relationship (Verbrugge, 1979). For instance, ex-fishermen organizations might be linked through friendship or family ties as well as through instrumental work relations. The multiplex exchange reduces transaction costs and allows to transfer incommensurable resources (Amarasinghe, 1989). As Heaney (2014) argues, multiplex networks are especially important when they mediate diverse processes, such as social mobilizations, that could not be explored or explained by a single network alone. It is further contented that multiplex networks provide more complete representations of the network structure (Marin and Hampton, 2007). For that purpose, we assembled a set of nine 29 × 29 adjacency matrices, where each matrix represents one type of inter-organizational binary directed ties:
(B) Information exchange,
(F) Political affinity,
(H) Perceived influence, and
Following Ernstson, Sörlin, and Elmqvist (2008), we aggregated these 9 matrices into a single matrix representing the multiplex network (N). Thus, the final matrix addition was N = A + B + C + D + E + F+ G + H + I. The resulting network has weighted directed ties with weights from zero to nine (0–9). For some calculations, we have symmetrized and/or dichotomized the network. In the case of family network, we used undirected ties given it is an undirected relationship. In the last step, we applied block modeling based on Euclidean distances (Wasserman and Faust, 1994; White, Boorman, and Breiger, 1976) to a set of binary uniplex networks as well as to an aggregated weighted directed multiplex network (see Figure 2a). According to conversion approach, we first analyzed uniplex networks separately, and then we continued with analysis of the multiplex network (Žiberna, 2014).
Block model is a simplified representation of a network that consists of groups of nodes that have similar relations to others (blocks or social positions) and patterns of relations among nodes and positions (social roles) (Faust and Wasserman, 1992; White, Boorman, and Breiger, 1976). Hence, the block modeling approach rests on the idea of structural equivalence where two nodes are structurally equivalent only if they are linked to the same set of nodes. Since structural equivalence can be very rarely found in real-world networks, structural similarity measures such as Euclidean distances are used. This way, the block modeling can succinctly summarize relational patterns within the network and thus capture its essential structural properties (e.g., Ernstson, Sörlin, and Elmqvist, 2008). We have used R 3.3.1 (R Core Team, 2013) and R packages igraph (Csardi and Nepusz, 2006), network (Butts, 2008a), sna (Butts, 2008b) for data processing, analysis, and visualization.
Cabo Pulmo social networks were composed of multiple kinds of relations, and multiple types of organizations (Figure 2a). The most common organizations that composed the network were: international nongovernmental organizations (NGOs), ex-fishermen organizations, universities and national NGOs and federal government agencies (see first column, Table 1). One third of the respondents of the survey were women (11), and half of the respondents (13) were directors of organizations.
Structural Properties of the Networks
We applied standard structural measures and performed conditional uniform graph (CUG) tests (Butts, 2008b) to determine whether the observed structural features significantly differ from what we would expect by chance. Density is a measure of social cohesion and overall network activity (Dietz, Ostrom, and Stern, 2003). In general, real-world social networks are sparse and have lower densities than random networks. This is also the case of all observed networks.
However, when comparing networks among themselves, cooperation and information exchange networks showed significantly higher values of density than the remaining ones (with the exception of the work network). By using quadratic assignment procedure, we found that there was a very strong correlation (r = 0.96) between the information exchange and cooperation matrices which suggests that information exchange was a key dimension of the cooperation. Cooperation and information exchange networks were further characterized through their inclusiveness expressed by a maximum value of connectedness (1); which means that there were no isolates in these networks.
On the other hand, very low density of political affinity and funding networks suggests that only a fraction of the organizations had financial and political ties. Reciprocity values were, rather surprisingly, lower than would be expected by chance in all 8 directed uniplex networks which goes against a general tendency to reciprocity (Wasserman and Faust, 1994). In contrast, transitivity values were higher than could be expected by chance with the exception of funding and political affinity networks. Both reciprocity and transitivity are measures associated with a non-hierarchical arrangement and clustering of the network. Based on that, we argue that organizations tend to constitute hierarchically organized cooperative clusters. This is supported by a higher density and inclusiveness of cooperation as well as information exchange networks that seemed to be crucial for the connectivity of the whole multiplex network. In contrast, conflict network is very sparse, moderately transitive, and does not substantially affect the multiplex structure. Degree centralization is a measure associated with hierarchy of the network (Freeman, 1978). We found that cooperation, information exchange, work, and friendship networks had higher values of centralization degree (Table 2). This would correspond with a situation where there are few active hubs that hold the network together (Bodin, Crona, and Ernstson, 2006). Besides the funding network, all uniplex networks are more degree centralized than we would expect by chance. Betweenness centrality then indicates the extent to which organizations lie among other organizations in the assemblage. In this sense, only few actors were in brokerage positions in information exchange, cooperation, and work networks (Table 2). CUG tests further shown that only cooperation and information exchange networks have higher betweenness centralization values than expected by chance. Friendship network, on the other hand, interestingly exhibited a lower level of betweenness than expected by chance. Since multiplex binary network resulted from an addition of adjacency matrices, it has the highest density as well as levels of reciprocity and transitivity (Table 2). This could be considered as an indication that multiplexity strengthens the tendency towards integration through establishment of group norms, cohesive support, and generalized exchange (Kivelä et al., 2014).
The multiplex network exhibited a structure with five positions (Figure 2b). The densest interactions were observed among blocks 1, 2, 3, and 5; conversely, block 4 was isolated from the rest. In general, we found blocks with heterogeneous composition but the same social role performance (White, Boorman, and Breiger, 1976). Based on the results of the block modeling and the knowledge of the case we got from the semi-structured interviews, we assigned tags to each block in order to explain how these structural roles come into being. Nevertheless, this classification does not imply that the agency of organizations is entirely constrained by its structural role (Emirbayer and Goodwin, 1994).
Block 1 was composed by a university, regional, national and international NGOs and one ex-fishermen organization. It is characterized by high internal density (0.76) and contained the most influential and strategist organizations in terms of their perceived influence; for that reason, we named them as “Visible leaders”. Based on the field research, we corroborated that members from Block 1 were key players to reach out to powerful actors in the conflict, including congressman, ministers as well as international media organizations. Block 1 had a strong bidirectional interaction with Block 2, “Scale-crossing brokers”, and received important inward interactions from Block 3, “Ecosystem managers” Block 2 was formed by a university and a national NGO whose territorial foci were regional and local. We named them “Scale-crossing brokers” given the strong reciprocate ties with Blocks 3 and 1. Organizations in Block 2 were involved in the creation, co-management and research of the marine protected area (MPA) since its beginnings. Members of Block 2 are known for their brokerage with donors and NGOs to sustain long-term human development and ecological research initiatives in the park. Thus, Block 2 members were able to upscale the local environmental conflict of Cabo Pulmo from local to international arenas through connections with “Visible leaders”. Besides, members from Block 2 acted also as the local intermediaries between Block 5, “Visionaries”, and Block 3, “Ecosystem managers”.
International and regional NGOs, National Park Office and ex-fishermen organizations were predominant in Block 3; all had been actively involved in the ecosystem management of CPNP. Most of the organizations in Block 3 have been in charge of the management plans, biodiversity monitoring, and law enforcement in the MPA. Block 4 was composed mostly of “Public sector” organizations: federal government agencies, developer and peripheral ex-fishermen organizations. Block 4 was structurally isolated from the rest due its low level of interactions. This block is the most heterogeneous in terms of composition and positions in the conflict, it contains organizations in favor, against and neutral position regarding the large scale development. In contrast, Block 5 was composed of universities, international NGOs and a donor, whom historically had promoted alternative visions for CPNP. We named them “Visionaries”, because some of them have been part of the environmental professionalized non-profit sector in Mexico since the nineties. Other organizations in this block were pioneers in the marine biology research of Cabo Pulmo's coral reef. Remarkably, Block 5 consisted of environmental international organizations whose main strategy is the legal action.
Cabo Pulmo is a key case for Mexico, in which multiple social networks facilitated a collective action that disrupted a policy process leading to a large-scale development and thus shielded the sustainable transformation of a marine protected area and enhanced the resilience of coastal ecosystems (Adger et al., 2005; Moore et al., 2014; Westley et al., 2013). Since 1995 Cabo Pulmo inhabitants and their partners in ecosystem management, started a transformation process towards sustainability, shifting from small-scale fishing towards a “new regime” of recreational ecotourism. Through time, they completed a full adaptive cycle across the four phases: exploitation, conservation, release and reorganization (Olsson, Galaz, and Boonstra, 2014). However, when Cabo Pulmo's ecosystem management organizations were struggling to institutionalize the new trajectory (Moore et al., 2014), they faced resistance from a large-scale touristic project.
In that moment, organizations from Blocks 2 and 3 who were working on ecosystem management at regional and local level, shifted their focus to the appropriate scale (national level) where the policy process of land-use decision was taking place (Ernstson, Sörlin, and Elmqvist, 2008). Cabo Pulmo's multiplex network distributed information and facilitated cooperation across scales to confront federal government consent to the development (Moore et al., 2014; Westley et al., 2013). “Scale-crossing brokers” upscale the local conflict through “Visible leaders” (Moore and Westley, 2011), who were already involved in other national and international struggles and had political and media contacts (Figure 2b) (Ernstson, Sörlin, and Elmqvist, 2008). “Visible leaders” used numerous tactics to affect decision making inside federal government: they persuaded policy-makers to act, provided information to them, changed international public opinion preferences and increased the salience of the issue.
The results suggest that structural positions of organizations in multiplex network allowed them to coordinate their actions and exchange resources effectively (Heaney, 2014). Further analyses further showed the crucial role of internally cohesive Blocks 1, 2, 3 and 5 (Figure 2b). Thus, information and resources were exchanged through four positions, each associated with a specific role. Our results contrast with those reported by Ernstson, Sörlin, and Elmqvist (2008) who found a social movement that protected a national park in Sweden. In this case, it was observed a core-periphery structure with three positions: core, semi-core and periphery. Core and semi-core organizations had many political contacts contrary to periphery associations which used intensively the park and had fewer political linkages. Block modeling revealed the complex decentralized social structure of Cabo Pulmo (Figure 2b). Such non-hierarchical configurations are associated with polycentric governance systems; innovation (Moore and Westley, 2011), social learning (Bodin, Crona, and Ernstson, 2006) and complex problem resolution in changing environments (Saxenian, 1996). The fact that only two organizations formed the Block 2 might suggest the vulnerable character of the decentralized social structure; the loss of “Scale-crossing brokers” in the system could provoke the fragmentation of the network. Structural analysis further shown statistically significant differences between the uniplex networks in Cabo Pulmo. While some social networks showed structural features associated with hierarchical configurations, such as cooperation and information exchange (Table 2), others as funding or political affinity were more heterarchical.
Multiplexity is an essential aspect of social relations, its main consequence is that ties in one network may have elicited the formation or dissolution of ties in other networks (Barbillon et al., 2016). In that sense the multiplexity perspective further exposed the interdependent web of relations in which Cabo Pulmo's organizations were embedded (Heaney, 2014). This view allowed us to identify the organizations who benefit from their position in the multilevel institutional context (Barbillon et al., 2016), and gave us information on how organizations exchange resources (information, work and funding) in the context of other preexistent ties such as friendship and family (Lazega and Pattison, 1999). Beyond the scope of the exploratory analysis presented here, we propose the utilization of Exponential Random Graph Models (ERGM) in order to assess the mechanisms that constrain status homophily in segmented multiplex networks (Robins et al., 2007).
Social network analysis provided a comprehensive analytical framework that gave us a complex insight into the underlying social structures that influenced land-use decision-making in coastal areas. However, one important limitation of the present study is that we did not obtain access to the high level decision-making sector and to the developers who promoted the large-scale development. As a result of this bias, the multiplex social network described in this paper is more representative of the collective action against the development.
Understanding the social processes that enable adaptive governance of social-ecological systems is essential for sustainability transformations and resilience of coastal ecosystems. Cabo Pulmo's multiplex network facilitated information and cooperation exchange among heterogeneous organizations. These linkages facilitated the collective action that derailed the policy process supporting the resort and thus contributed crucially to the cancellation of this large-scale development (Moore et al., 2014). Cabo Pulmo's multiplex network has overcome sociopolitical barriers by connecting different sectors, administrative jurisdictions and spatial scales in order to protect ecosystem services and livelihoods (Folke et al., 2005). The theoretical relevance of this study is that it explores how multiple social networks disrupted an unwanted policy. The decision of canceling the environmental permit for building the large-scale resort promotes the provision of ecosystem services from Cabo Pulmo National Park with a net value of 1.1 million of dollars per year (Reyes-Bonilla et al., 2014). Thus, this case has provided evidence of the role of social networks in the resilience of social-ecological systems.
We would like to thank all the organizations that answered the survey. M. Luisa Martínez, Harlan Koff and Martin Hilbert provided fundamental suggestions that improved early versions of this manuscript. Ellen Dean, Miguel Equihua Zamora and Bruce Cronin facilitated Alfonso Langle's training on methods abroad. Ernesto Isunza suggested novel ways to formulate the survey. Juan José Von Thaden kindly prepared the map. Tomáš Diviák provided helpful comments on application of block modeling. A. Langle would like to thank the Mexican Council of Science and Technology (CONACYT) for Doctoral scholarship number 279962. Petr Ocelík's research activities are financed by “Europe in a Changing International Environment II” (MUNI/A/1113/2015) research project.