The relationship between institutions interventions and municipalities is presented in Fig. 4. Note that FAO, USAID, Inter-American Development Bank (BID), the Spanish Agency for International Development Cooperation (AECID), and the World Food Program (PMA) are international cooperation with interventions reported in the 153 municipalities, therefore they are not shown in this figure. Additionally, public institutions have by law interventions in all the country, however, some of these institutions as the SAG and Forest Conservation Institute (ICF) have institutional regionals, but limited to 8 and 12 regionals, respectively. Because the response to climate change in the agricultural sector requires a close coordination with policies to reduce deforestation, protect biodiversity and manage water resources Lennox, 2012 [28] the actualization of the Honduras’ Nationally Determined Contributions presented in the COP26 [29] have 2 common objectives between the 2 institutions. This was done since the links of land use change and forestry sector are key to cope climate change, and to interrelate the regionals of both institutions to cover more regions. This is not unique for Honduras or the region, in Ethiopia, Bergesten (2019) [30] observed a collaboration gap around forest and wildlife conservation, but dense collaboration around agricultural production, defining this as a collaborative misfit when the governance actors are not linked even working in the same ecological issue.
Through Fig. 4, it is possible to visualize patterns of fat tails especially in relationships that are in the peripheral areas of the network. This pattern reveals an important dependence in two senses: a) many municipalities are interacting with few institutional actors and b) many institutions are linked to one or few municipalities. In the case of a) it would imply a structure of specialization in care at the territorial level, but it could also express egocentric relationships with a high level of risk and fragility. Case b) would imply a territory with a high demand for attention or concentration of actions, a process of over-presence of institutional actors in the same region could also be configured.
Figure 4. Institutions and municipalities connections, as a social-network graph (a) and specific features of it described in the map (b). In the map the study area is shaded in light gray, in darkest gray for four ego networks, in red for the departmental capitals in the network core, including the Central District, and in orange for the three isolated networks. Black lines stand for the main roads and red lines for the unpaved roads. The numbers are related between both figures.
Municipalities in which few institutions do their interventions are situated in the periphery of the network graph (number 1–3 and 5 in Fig. 4) and located close to the frontiers. The ego network 1 and 2, near to El Salvador have in fact their own Municipalities Association (Mancomunidad in Spanish) as their unique institution connected to the municipalities. Thus, these municipalities unify efforts to common strategies plans. Moreover, these two ego networks are connected to themselves through one of their municipalities to the core network. In the case of the tail assigned with 1, the association includes the municipality of Marcala (1208) in which the DenomOrig (see section 3.1) is located and is linked to 10 institutions. This private project connects 19 municipalities (in and out the study area) and more than 2,300 coffee producers (http://www.docafemarcala.org/). In the case of the tail assigned with 2, Guajiquiro municipality (1206) is the one that connect the other municipalities with the core and to the DenomOrig.
Ego networks 3 and 5 are indeed connected to the network core by their Municipality Association. In the case of tail assigned with 3, this is connected by the international Amigos de la Tierra (https://www.tierra.org/honduras/), and for tail-5 near to Guatemala by the Mennonite Social Action Committee (CASM in Spanish), acting as a bridging actor with international cooperation and donors as OXFAM, Lutheran World Relief, Help in Action (Ayuda en Acción), among others (https://casm.hn/proyectos). By analyzing the multi-level water governance network in Central America, Hileman and Lubell (2018) [31] discussed some opportunities and constraints within the network. Analogous to the international cooperation bridging actors in Fig. 5, they observed that fostering cooperation can help as a bridge to regional actors, however it makes the network vulnerable to the exit problem, when a project and/or its funding ends. This could be more evident for the insolated networks 7 and 8 depicted in Fig. 5, which can be part of international cooperation projects (i.e., USAID, FAO), but somehow, they are disconnected from the institutional-municipality network obtained in this study. We cannot say the same for the isolated network assigned with 4, since these three municipalities are close to the capital. In fact, Valle de Ángeles (826) municipality is a touristic place and a dormitory municipality for people working in the capital.
Conversely, 34 institutions are linked to the Central District (801; tail-6 in Fig. 4). This is to be expected because it is where government secretariats are located. Figure 5 also shows that 8 departmental capitals (red municipalities in map), including the Central District are part of the main network core (border in red square in Fig. 4). In addition to the municipality of Intibucá which is separated to La Esperanza (departmental capital) just by a street. Hence, they are highly connected. In fact, among the municipalities with the large in-degree (> 4%) related to the number of institutions linked to them are these capitals, Intibucá and Marcala.
Municipalities outside the network core and not discussed in the four periphery networks, are dispersed -mostly- between the development regions 3 and 14 (see Fig. 1). Both regions possess the largest multidimensional poverty incidence (> 85%), which measures the deprivation of a person of more than 50% of the weighted sum of indicators related to heath, education, work and living conditions between 2012 and 2016 [32]. Indeed, in terms of physical connections among the study area, these municipalities are only connected by unpaved roads (red in map, Fig. 4) through a highly mountainous region. The country's main road network (black in map, Fig. 4) mainly connects the departmental capitals and in the south and west, the connection with the borders is also observed.
The deficiency in the national highway network is one of the multi-causes for the limited intervention of some institutions. Anderson and Laerhoven (2007) [33], evaluate how rural local government representatives in 390 municipal governments in Brazil, Chile, Mexico and Peru, relate and involve farmers in planning, implementation, and monitoring of public services in the agricultural sector. They observed that local politicians were less interested in participatory governance, at least in Peru. The authors hypothesize that the greater prevalence of poverty and socioeconomic inequalities in Peruvian rural societies, compared to the other three countries, but like Honduran rural, could be related to the lack of participatory governance. This is relevant because the potentiality of local actors and resources are needed for endogenous and bottom-up development (Bosworth et al. 2020) [34].
The reader needs to consider that there are other links that are not shown in Fig. 4, as the connections with municipalities from other countries. This applies for municipalities from the development region 3 and 14. The Trifinio Plan is a regional organization that is part of the Central American Integration System (SICA), which seeks to develop a process for managing the environment and the territory, in order for it to become the possibility of improving conditions of life in 8 municipalities from El Salvador, 15 in Guatemala and 22 in Honduras (https://www.plantrifinio.int/). Hence, these municipalities in Honduras can also present regional interventions and were not informed in the survey.