Local Collaboration and Influence in Natural Resource Management in Amani Nature Reserve in Tanzania: A Cross-Sectional Social Network Analysis

Abstract

Social network analysis (SNA) in natural resource management has received muchattention in recent literature, where different characteristics of social networks havebeen revealed to serve various functions to demonstrate effective collaboration in thegovernance of natural resources. In this paper, we consider a Bayesian model selectionapproach for exponential random graph models (ERGMs), which is based on theposterior distribution. We use a combination of Bayesian and MCMC techniques,which circumvents the need to calculate the normalising constants. This method isillustrated using a cross-sectional network of five types of relational ties and influenceas an actor ' s attribute in the context of local collaboration practices of natural resourcemanagement in two rural communities of Amani Nature Reserve in Tanzania. Basedon the positive values of parameters related to network structural characteristics inthe competing models, our results provide an overall understanding of actorsperceived as influential, and the contributions of their attributes and networkstructures in explaining influence nominations.  Keywords: actor attributes, Bayesian model selection, Bayes factors, exponential randomgraph models, local collaborative practices of natural resource management, network ties,social networks